diff --git a/en/resources/responses/index.md b/en/resources/responses/index.md index c2a25b04..a78b136e 100644 --- a/en/resources/responses/index.md +++ b/en/resources/responses/index.md @@ -24,7 +24,7 @@ the `background` parameter set to `true` can be cancelled. Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -76,7 +76,7 @@ the `background` parameter set to `true` can be cancelled. A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -88,7 +88,7 @@ the `background` parameter set to `true` can be cancelled. - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -174,11 +174,11 @@ the `background` parameter set to `true` can be cancelled. - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -216,7 +216,7 @@ the `background` parameter set to `true` can be cancelled. The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -251,7 +251,7 @@ the `background` parameter set to `true` can be cancelled. - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -484,7 +484,7 @@ the `background` parameter set to `true` can be cancelled. - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -528,11 +528,11 @@ the `background` parameter set to `true` can be cancelled. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -583,11 +583,11 @@ the `background` parameter set to `true` can be cancelled. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -644,7 +644,7 @@ the `background` parameter set to `true` can be cancelled. The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -652,7 +652,7 @@ the `background` parameter set to `true` can be cancelled. A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -699,7 +699,7 @@ the `background` parameter set to `true` can be cancelled. - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -735,7 +735,7 @@ the `background` parameter set to `true` can be cancelled. The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -775,7 +775,7 @@ the `background` parameter set to `true` can be cancelled. The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -877,11 +877,11 @@ the `background` parameter set to `true` can be cancelled. - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -889,15 +889,15 @@ the `background` parameter set to `true` can be cancelled. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -963,7 +963,7 @@ the `background` parameter set to `true` can be cancelled. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -1030,7 +1030,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -1099,7 +1099,7 @@ the `background` parameter set to `true` can be cancelled. - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -1119,7 +1119,7 @@ the `background` parameter set to `true` can be cancelled. - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -1131,15 +1131,15 @@ the `background` parameter set to `true` can be cancelled. - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -1169,23 +1169,23 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -1201,11 +1201,11 @@ the `background` parameter set to `true` can be cancelled. - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -1229,15 +1229,15 @@ the `background` parameter set to `true` can be cancelled. - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -1259,11 +1259,11 @@ the `background` parameter set to `true` can be cancelled. - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -1275,7 +1275,7 @@ the `background` parameter set to `true` can be cancelled. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -1299,11 +1299,11 @@ the `background` parameter set to `true` can be cancelled. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -1313,7 +1313,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -1325,11 +1325,11 @@ the `background` parameter set to `true` can be cancelled. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -1347,7 +1347,7 @@ the `background` parameter set to `true` can be cancelled. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -1513,11 +1513,11 @@ the `background` parameter set to `true` can be cancelled. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -1534,23 +1534,23 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -1575,7 +1575,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -1583,7 +1583,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -1648,12 +1648,12 @@ the `background` parameter set to `true` can be cancelled. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -1743,7 +1743,7 @@ the `background` parameter set to `true` can be cancelled. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -1801,7 +1801,7 @@ the `background` parameter set to `true` can be cancelled. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -1848,7 +1848,7 @@ the `background` parameter set to `true` can be cancelled. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -1963,7 +1963,7 @@ the `background` parameter set to `true` can be cancelled. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -1971,7 +1971,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -1985,7 +1985,7 @@ the `background` parameter set to `true` can be cancelled. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -2111,7 +2111,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2187,7 +2187,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2199,15 +2199,15 @@ the `background` parameter set to `true` can be cancelled. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -2225,7 +2225,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2261,7 +2261,7 @@ the `background` parameter set to `true` can be cancelled. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -2273,7 +2273,7 @@ the `background` parameter set to `true` can be cancelled. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -2305,7 +2305,7 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -2315,19 +2315,19 @@ the `background` parameter set to `true` can be cancelled. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -2341,7 +2341,7 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2355,11 +2355,11 @@ the `background` parameter set to `true` can be cancelled. - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -2371,7 +2371,7 @@ the `background` parameter set to `true` can be cancelled. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -2401,11 +2401,11 @@ the `background` parameter set to `true` can be cancelled. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -2415,7 +2415,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2427,11 +2427,11 @@ the `background` parameter set to `true` can be cancelled. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -2449,7 +2449,7 @@ the `background` parameter set to `true` can be cancelled. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -2548,11 +2548,11 @@ the `background` parameter set to `true` can be cancelled. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -2569,23 +2569,23 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -2610,7 +2610,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2618,7 +2618,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -2683,12 +2683,12 @@ the `background` parameter set to `true` can be cancelled. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -2778,7 +2778,7 @@ the `background` parameter set to `true` can be cancelled. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -2804,7 +2804,7 @@ the `background` parameter set to `true` can be cancelled. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2851,7 +2851,7 @@ the `background` parameter set to `true` can be cancelled. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -2966,7 +2966,7 @@ the `background` parameter set to `true` can be cancelled. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -2974,7 +2974,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -2996,7 +2996,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -3040,7 +3040,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -3052,15 +3052,15 @@ the `background` parameter set to `true` can be cancelled. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -3078,7 +3078,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -3114,7 +3114,7 @@ the `background` parameter set to `true` can be cancelled. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -3126,7 +3126,7 @@ the `background` parameter set to `true` can be cancelled. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -3158,7 +3158,7 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -3168,19 +3168,19 @@ the `background` parameter set to `true` can be cancelled. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -3194,7 +3194,7 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -3208,7 +3208,7 @@ the `background` parameter set to `true` can be cancelled. - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -3257,7 +3257,7 @@ the `background` parameter set to `true` can be cancelled. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -3288,7 +3288,7 @@ the `background` parameter set to `true` can be cancelled. - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -3300,7 +3300,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -3330,7 +3330,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -3338,7 +3338,7 @@ the `background` parameter set to `true` can be cancelled. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -3417,15 +3417,15 @@ the `background` parameter set to `true` can be cancelled. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -3467,7 +3467,7 @@ the `background` parameter set to `true` can be cancelled. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -3489,11 +3489,11 @@ the `background` parameter set to `true` can be cancelled. Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -3507,11 +3507,11 @@ the `background` parameter set to `true` can be cancelled. - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -3535,7 +3535,7 @@ the `background` parameter set to `true` can be cancelled. - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -3543,7 +3543,7 @@ the `background` parameter set to `true` can be cancelled. - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -3607,11 +3607,11 @@ the `background` parameter set to `true` can be cancelled. - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -3635,11 +3635,11 @@ the `background` parameter set to `true` can be cancelled. - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -3725,11 +3725,11 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -3775,11 +3775,11 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -3803,7 +3803,7 @@ the `background` parameter set to `true` can be cancelled. - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -3831,11 +3831,11 @@ the `background` parameter set to `true` can be cancelled. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -3845,7 +3845,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -3893,11 +3893,11 @@ the `background` parameter set to `true` can be cancelled. - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -3927,16 +3927,16 @@ the `background` parameter set to `true` can be cancelled. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -3997,7 +3997,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -4047,7 +4047,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -4089,7 +4089,7 @@ the `background` parameter set to `true` can be cancelled. The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -4147,7 +4147,7 @@ the `background` parameter set to `true` can be cancelled. - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -4414,11 +4414,11 @@ the `background` parameter set to `true` can be cancelled. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -4475,7 +4475,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -4562,7 +4562,7 @@ the `background` parameter set to `true` can be cancelled. - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -4654,7 +4654,7 @@ the `background` parameter set to `true` can be cancelled. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -4715,11 +4715,11 @@ the `background` parameter set to `true` can be cancelled. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -4791,11 +4791,11 @@ the `background` parameter set to `true` can be cancelled. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -4846,7 +4846,7 @@ the `background` parameter set to `true` can be cancelled. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -4925,7 +4925,7 @@ the `background` parameter set to `true` can be cancelled. Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -4963,7 +4963,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -4997,11 +4997,11 @@ the `background` parameter set to `true` can be cancelled. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -5011,7 +5011,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5023,11 +5023,11 @@ the `background` parameter set to `true` can be cancelled. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -5045,7 +5045,7 @@ the `background` parameter set to `true` can be cancelled. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -5144,11 +5144,11 @@ the `background` parameter set to `true` can be cancelled. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -5165,23 +5165,23 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -5206,7 +5206,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5214,7 +5214,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -5279,12 +5279,12 @@ the `background` parameter set to `true` can be cancelled. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -5374,7 +5374,7 @@ the `background` parameter set to `true` can be cancelled. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -5400,7 +5400,7 @@ the `background` parameter set to `true` can be cancelled. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5447,7 +5447,7 @@ the `background` parameter set to `true` can be cancelled. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -5562,7 +5562,7 @@ the `background` parameter set to `true` can be cancelled. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5570,7 +5570,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -5592,7 +5592,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5636,7 +5636,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5648,15 +5648,15 @@ the `background` parameter set to `true` can be cancelled. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -5674,7 +5674,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5710,7 +5710,7 @@ the `background` parameter set to `true` can be cancelled. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -5722,7 +5722,7 @@ the `background` parameter set to `true` can be cancelled. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -5754,7 +5754,7 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -5764,19 +5764,19 @@ the `background` parameter set to `true` can be cancelled. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -5790,7 +5790,7 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5846,11 +5846,11 @@ the `background` parameter set to `true` can be cancelled. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -5860,7 +5860,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -5872,11 +5872,11 @@ the `background` parameter set to `true` can be cancelled. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -5894,7 +5894,7 @@ the `background` parameter set to `true` can be cancelled. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -5993,11 +5993,11 @@ the `background` parameter set to `true` can be cancelled. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -6014,23 +6014,23 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -6055,7 +6055,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -6063,7 +6063,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -6128,12 +6128,12 @@ the `background` parameter set to `true` can be cancelled. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -6223,7 +6223,7 @@ the `background` parameter set to `true` can be cancelled. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -6249,7 +6249,7 @@ the `background` parameter set to `true` can be cancelled. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -6296,7 +6296,7 @@ the `background` parameter set to `true` can be cancelled. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -6411,7 +6411,7 @@ the `background` parameter set to `true` can be cancelled. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -6419,7 +6419,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -6441,7 +6441,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -6485,7 +6485,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -6497,15 +6497,15 @@ the `background` parameter set to `true` can be cancelled. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -6523,7 +6523,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -6559,7 +6559,7 @@ the `background` parameter set to `true` can be cancelled. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -6571,7 +6571,7 @@ the `background` parameter set to `true` can be cancelled. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -6603,7 +6603,7 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -6613,19 +6613,19 @@ the `background` parameter set to `true` can be cancelled. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -6639,7 +6639,7 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -6683,7 +6683,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -6713,7 +6713,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -6721,7 +6721,7 @@ the `background` parameter set to `true` can be cancelled. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -6800,15 +6800,15 @@ the `background` parameter set to `true` can be cancelled. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -6850,7 +6850,7 @@ the `background` parameter set to `true` can be cancelled. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -6874,11 +6874,11 @@ the `background` parameter set to `true` can be cancelled. - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -6886,7 +6886,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -6928,7 +6928,7 @@ the `background` parameter set to `true` can be cancelled. - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -6964,7 +6964,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -7028,7 +7028,7 @@ the `background` parameter set to `true` can be cancelled. - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -7132,7 +7132,7 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -7182,7 +7182,7 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -7206,7 +7206,7 @@ the `background` parameter set to `true` can be cancelled. The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -7236,16 +7236,16 @@ the `background` parameter set to `true` can be cancelled. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -7287,11 +7287,11 @@ the `background` parameter set to `true` can be cancelled. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -7301,7 +7301,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -7353,7 +7353,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -7383,7 +7383,7 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -7459,7 +7459,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -7491,7 +7491,7 @@ the `background` parameter set to `true` can be cancelled. Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -7620,7 +7620,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -7694,11 +7694,11 @@ the `background` parameter set to `true` can be cancelled. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -7708,7 +7708,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -7720,11 +7720,11 @@ the `background` parameter set to `true` can be cancelled. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -7742,7 +7742,7 @@ the `background` parameter set to `true` can be cancelled. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -7841,11 +7841,11 @@ the `background` parameter set to `true` can be cancelled. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -7862,23 +7862,23 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -7903,7 +7903,7 @@ the `background` parameter set to `true` can be cancelled. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -7911,7 +7911,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -7976,12 +7976,12 @@ the `background` parameter set to `true` can be cancelled. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -8071,7 +8071,7 @@ the `background` parameter set to `true` can be cancelled. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -8097,7 +8097,7 @@ the `background` parameter set to `true` can be cancelled. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -8144,7 +8144,7 @@ the `background` parameter set to `true` can be cancelled. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -8259,7 +8259,7 @@ the `background` parameter set to `true` can be cancelled. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -8267,7 +8267,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -8289,7 +8289,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -8333,7 +8333,7 @@ the `background` parameter set to `true` can be cancelled. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -8345,15 +8345,15 @@ the `background` parameter set to `true` can be cancelled. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -8371,7 +8371,7 @@ the `background` parameter set to `true` can be cancelled. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -8407,7 +8407,7 @@ the `background` parameter set to `true` can be cancelled. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -8419,7 +8419,7 @@ the `background` parameter set to `true` can be cancelled. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -8451,7 +8451,7 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -8461,19 +8461,19 @@ the `background` parameter set to `true` can be cancelled. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -8487,7 +8487,7 @@ the `background` parameter set to `true` can be cancelled. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -8495,7 +8495,7 @@ the `background` parameter set to `true` can be cancelled. - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -8504,17 +8504,17 @@ the `background` parameter set to `true` can be cancelled. We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -8522,15 +8522,15 @@ the `background` parameter set to `true` can be cancelled. The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -8646,19 +8646,19 @@ the `background` parameter set to `true` can be cancelled. - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -8667,7 +8667,7 @@ the `background` parameter set to `true` can be cancelled. The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -8687,11 +8687,11 @@ the `background` parameter set to `true` can be cancelled. A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -8713,7 +8713,7 @@ the `background` parameter set to `true` can be cancelled. - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -8732,14 +8732,14 @@ the `background` parameter set to `true` can be cancelled. - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -8755,7 +8755,7 @@ the `background` parameter set to `true` can be cancelled. - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -8779,7 +8779,7 @@ the `background` parameter set to `true` can be cancelled. - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -8811,7 +8811,7 @@ the `background` parameter set to `true` can be cancelled. - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -8825,12 +8825,12 @@ the `background` parameter set to `true` can be cancelled. - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -8931,7 +8931,7 @@ the `background` parameter set to `true` can be cancelled. A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -8952,7 +8952,7 @@ the `background` parameter set to `true` can be cancelled. - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -8965,14 +8965,14 @@ the `background` parameter set to `true` can be cancelled. - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -9281,7 +9281,7 @@ Learn when and how to compact long-running conversations in the [conversation st ### Body Parameters -- `model: "gpt-5.6-sol" or "gpt-5.6-terra" or "gpt-5.6-luna" or 92 more or string` +- `model: "gpt-5.6-sol" or "gpt-5.6-terra" or "gpt-5.6-luna" or 92 more or string or null` Model ID used to generate the response, like `gpt-5` or `o3`. OpenAI offers a wide range of models with different capabilities, performance characteristics, and price points. Refer to the [model guide](/docs/models) to browse and compare available models. @@ -9481,7 +9481,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `string` -- `input: optional string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` +- `input: optional string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` Text, image, or file inputs to the model, used to generate a response @@ -9561,11 +9561,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -9603,7 +9603,7 @@ Learn when and how to compact long-running conversations in the [conversation st The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -9638,7 +9638,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -9871,7 +9871,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -9915,11 +9915,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -9970,11 +9970,11 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -10031,7 +10031,7 @@ Learn when and how to compact long-running conversations in the [conversation st The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -10039,7 +10039,7 @@ Learn when and how to compact long-running conversations in the [conversation st A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -10086,7 +10086,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -10122,7 +10122,7 @@ Learn when and how to compact long-running conversations in the [conversation st The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -10162,7 +10162,7 @@ Learn when and how to compact long-running conversations in the [conversation st The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -10264,11 +10264,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -10276,15 +10276,15 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -10350,7 +10350,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -10417,7 +10417,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -10486,7 +10486,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -10506,7 +10506,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -10518,15 +10518,15 @@ Learn when and how to compact long-running conversations in the [conversation st - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -10556,23 +10556,23 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -10588,11 +10588,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -10616,15 +10616,15 @@ Learn when and how to compact long-running conversations in the [conversation st - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -10646,11 +10646,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -10662,7 +10662,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -10686,11 +10686,11 @@ Learn when and how to compact long-running conversations in the [conversation st The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -10700,7 +10700,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -10712,11 +10712,11 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -10734,7 +10734,7 @@ Learn when and how to compact long-running conversations in the [conversation st The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -10900,11 +10900,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -10921,23 +10921,23 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -10962,7 +10962,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -10970,7 +10970,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -11035,12 +11035,12 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -11130,7 +11130,7 @@ Learn when and how to compact long-running conversations in the [conversation st An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -11188,7 +11188,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -11235,7 +11235,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -11350,7 +11350,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -11358,7 +11358,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -11372,7 +11372,7 @@ Learn when and how to compact long-running conversations in the [conversation st An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -11498,7 +11498,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -11574,7 +11574,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -11586,15 +11586,15 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -11612,7 +11612,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -11648,7 +11648,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -11660,7 +11660,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -11692,7 +11692,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -11702,19 +11702,19 @@ Learn when and how to compact long-running conversations in the [conversation st - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -11728,7 +11728,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -11742,11 +11742,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -11758,7 +11758,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -11788,11 +11788,11 @@ Learn when and how to compact long-running conversations in the [conversation st The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -11802,7 +11802,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -11814,11 +11814,11 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -11836,7 +11836,7 @@ Learn when and how to compact long-running conversations in the [conversation st The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -11935,11 +11935,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -11956,23 +11956,23 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -11997,7 +11997,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -12005,7 +12005,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -12070,12 +12070,12 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -12165,7 +12165,7 @@ Learn when and how to compact long-running conversations in the [conversation st An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -12191,7 +12191,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -12238,7 +12238,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -12353,7 +12353,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -12361,7 +12361,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -12383,7 +12383,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -12427,7 +12427,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -12439,15 +12439,15 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -12465,7 +12465,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -12501,7 +12501,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -12513,7 +12513,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -12545,7 +12545,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -12555,19 +12555,19 @@ Learn when and how to compact long-running conversations in the [conversation st - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -12581,7 +12581,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -12595,7 +12595,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -12644,7 +12644,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -12675,7 +12675,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -12687,7 +12687,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -12717,7 +12717,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -12725,7 +12725,7 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -12804,15 +12804,15 @@ Learn when and how to compact long-running conversations in the [conversation st - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -12854,7 +12854,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -12876,11 +12876,11 @@ Learn when and how to compact long-running conversations in the [conversation st Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -12894,11 +12894,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -12922,7 +12922,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -12930,7 +12930,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -12994,11 +12994,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -13022,11 +13022,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -13112,11 +13112,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -13162,11 +13162,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -13190,7 +13190,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -13218,11 +13218,11 @@ Learn when and how to compact long-running conversations in the [conversation st The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -13232,7 +13232,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -13280,11 +13280,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -13314,16 +13314,16 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -13384,7 +13384,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -13434,7 +13434,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -13476,7 +13476,7 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -13534,20 +13534,20 @@ Learn when and how to compact long-running conversations in the [conversation st - `"program_output"` -- `instructions: optional string` +- `instructions: optional string or null` A system (or developer) message inserted into the model's context. When used along with `previous_response_id`, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses. -- `previous_response_id: optional string` +- `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. -- `prompt_cache_key: optional string` +- `prompt_cache_key: optional string or null` A key to use when reading from or writing to the prompt cache. -- `prompt_cache_options: optional object { mode, ttl }` +- `prompt_cache_options: optional object { mode, ttl } or null` Options for prompt caching. Supported for `gpt-5.6` and later models. By default, OpenAI automatically chooses one implicit cache breakpoint. You can add explicit breakpoints to content blocks with `prompt_cache_breakpoint`. Each request can write up to four breakpoints. For cache matching, OpenAI considers up to the latest 80 breakpoints in the conversation, without a content-block lookback limit. Set `mode` to `explicit` to disable the implicit breakpoint. The `ttl` defaults to `30m`, which is currently the only supported value. See the [prompt caching guide](/docs/guides/prompt-caching) for current details. @@ -13565,7 +13565,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"30m"` -- `prompt_cache_retention: optional "in_memory" or "24h"` +- `prompt_cache_retention: optional "in_memory" or "24h" or null` How long to retain a prompt cache entry created by this request. @@ -13573,7 +13573,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"24h"` -- `service_tier: optional "auto" or "default" or "fast" or 2 more` +- `service_tier: optional "auto" or "default" or "fast" or 2 more or null` Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to '[flex](/docs/guides/flex-processing)', then the request will be processed with the Flex Processing service tier. - To opt-in to [Fast mode](/api/docs/guides/fast-mode) at the request level, include the `service_tier=fast` or `service_tier=priority` parameter for Responses or Chat Completions. The response will show `service_tier=priority` regardless of if you specify `service_tier=fast` or `priority` in your request. - When not set, the default behavior is 'auto'. When the `service_tier` parameter is set, the response body will include the `service_tier` value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter. @@ -13850,11 +13850,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -13884,11 +13884,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"original"` - - `file_id: string` + - `file_id: string or null` The identifier of an uploaded file that contains the screenshot. - - `image_url: string` + - `image_url: string or null` The URL of the screenshot image. @@ -13932,7 +13932,7 @@ Learn when and how to compact long-running conversations in the [conversation st The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -13990,7 +13990,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend phase on all assistant messages — dropping it can degrade performance. Not used for user messages. @@ -14077,7 +14077,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -14122,7 +14122,7 @@ Learn when and how to compact long-running conversations in the [conversation st Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -14160,7 +14160,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -14194,11 +14194,11 @@ Learn when and how to compact long-running conversations in the [conversation st The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -14208,7 +14208,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -14220,11 +14220,11 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -14242,7 +14242,7 @@ Learn when and how to compact long-running conversations in the [conversation st The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -14408,11 +14408,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -14429,23 +14429,23 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -14470,7 +14470,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -14478,7 +14478,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -14543,12 +14543,12 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -14638,7 +14638,7 @@ Learn when and how to compact long-running conversations in the [conversation st An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -14696,7 +14696,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -14743,7 +14743,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -14858,7 +14858,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -14866,7 +14866,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -14880,7 +14880,7 @@ Learn when and how to compact long-running conversations in the [conversation st An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -15006,7 +15006,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15082,7 +15082,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15094,15 +15094,15 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -15120,7 +15120,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15156,7 +15156,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -15168,7 +15168,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -15200,7 +15200,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -15210,19 +15210,19 @@ Learn when and how to compact long-running conversations in the [conversation st - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -15236,7 +15236,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15292,11 +15292,11 @@ Learn when and how to compact long-running conversations in the [conversation st The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -15306,7 +15306,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15318,11 +15318,11 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -15340,7 +15340,7 @@ Learn when and how to compact long-running conversations in the [conversation st The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -15439,11 +15439,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -15460,23 +15460,23 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -15501,7 +15501,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15509,7 +15509,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -15574,12 +15574,12 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -15669,7 +15669,7 @@ Learn when and how to compact long-running conversations in the [conversation st An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -15695,7 +15695,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15742,7 +15742,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -15857,7 +15857,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15865,7 +15865,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -15887,7 +15887,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15931,7 +15931,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -15943,15 +15943,15 @@ Learn when and how to compact long-running conversations in the [conversation st Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -15969,7 +15969,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -16005,7 +16005,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -16017,7 +16017,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -16049,7 +16049,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -16059,19 +16059,19 @@ Learn when and how to compact long-running conversations in the [conversation st - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -16085,7 +16085,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -16143,7 +16143,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -16220,11 +16220,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -16310,7 +16310,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -16358,7 +16358,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -16401,11 +16401,11 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -16462,7 +16462,7 @@ Learn when and how to compact long-running conversations in the [conversation st The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -16470,7 +16470,7 @@ Learn when and how to compact long-running conversations in the [conversation st A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -16517,7 +16517,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -16553,7 +16553,7 @@ Learn when and how to compact long-running conversations in the [conversation st The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -16593,7 +16593,7 @@ Learn when and how to compact long-running conversations in the [conversation st The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -16719,11 +16719,11 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -16774,7 +16774,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -16821,7 +16821,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -16829,7 +16829,7 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -16908,15 +16908,15 @@ Learn when and how to compact long-running conversations in the [conversation st - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -16958,7 +16958,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -16982,11 +16982,11 @@ Learn when and how to compact long-running conversations in the [conversation st - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -16994,7 +16994,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -17036,7 +17036,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -17072,7 +17072,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -17136,7 +17136,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -17240,7 +17240,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -17290,7 +17290,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -17314,7 +17314,7 @@ Learn when and how to compact long-running conversations in the [conversation st The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -17342,11 +17342,11 @@ Learn when and how to compact long-running conversations in the [conversation st The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -17356,7 +17356,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -17408,7 +17408,7 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -17438,16 +17438,16 @@ Learn when and how to compact long-running conversations in the [conversation st - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -17491,7 +17491,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -17558,7 +17558,7 @@ Learn when and how to compact long-running conversations in the [conversation st The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -17757,12 +17757,12 @@ as input for the model's response. ### Body Parameters -- `background: optional boolean` +- `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). -- `context_management: optional array of object { type, compact_threshold }` +- `context_management: optional array of object { type, compact_threshold } or null` Context management configuration for this request. @@ -17770,11 +17770,11 @@ as input for the model's response. The context management entry type. Currently only 'compaction' is supported. - - `compact_threshold: optional number` + - `compact_threshold: optional number or null` Token threshold at which compaction should be triggered for this entry. -- `conversation: optional string or ResponseConversationParam` +- `conversation: optional string or ResponseConversationParam or null` The conversation that this response belongs to. Items from this conversation are prepended to `input_items` for this response request. Input items and output items from this response are automatically added to this conversation after this response completes. @@ -17791,7 +17791,7 @@ as input for the model's response. The unique ID of the conversation. -- `include: optional array of ResponseIncludable` +- `include: optional array of ResponseIncludable or null` Specify additional output data to include in the model response. Currently supported values are: @@ -17909,11 +17909,11 @@ as input for the model's response. - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -17951,7 +17951,7 @@ as input for the model's response. The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -17986,7 +17986,7 @@ as input for the model's response. - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -18219,7 +18219,7 @@ as input for the model's response. - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -18263,11 +18263,11 @@ as input for the model's response. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -18318,11 +18318,11 @@ as input for the model's response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -18379,7 +18379,7 @@ as input for the model's response. The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -18387,7 +18387,7 @@ as input for the model's response. A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -18434,7 +18434,7 @@ as input for the model's response. - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -18470,7 +18470,7 @@ as input for the model's response. The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -18510,7 +18510,7 @@ as input for the model's response. The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -18612,11 +18612,11 @@ as input for the model's response. - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -18624,15 +18624,15 @@ as input for the model's response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -18698,7 +18698,7 @@ as input for the model's response. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -18765,7 +18765,7 @@ as input for the model's response. The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -18834,7 +18834,7 @@ as input for the model's response. - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -18854,7 +18854,7 @@ as input for the model's response. - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -18866,15 +18866,15 @@ as input for the model's response. - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -18904,23 +18904,23 @@ as input for the model's response. - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -18936,11 +18936,11 @@ as input for the model's response. - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -18964,15 +18964,15 @@ as input for the model's response. - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -18994,11 +18994,11 @@ as input for the model's response. - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -19010,7 +19010,7 @@ as input for the model's response. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -19034,11 +19034,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -19048,7 +19048,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -19060,11 +19060,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -19082,7 +19082,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -19248,11 +19248,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -19269,23 +19269,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -19310,7 +19310,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -19318,7 +19318,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -19383,12 +19383,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -19478,7 +19478,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -19536,7 +19536,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -19583,7 +19583,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -19698,7 +19698,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -19706,7 +19706,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -19720,7 +19720,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -19846,7 +19846,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -19922,7 +19922,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -19934,15 +19934,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -19960,7 +19960,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -19996,7 +19996,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -20008,7 +20008,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -20040,7 +20040,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -20050,19 +20050,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -20076,7 +20076,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20090,11 +20090,11 @@ as input for the model's response. - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -20106,7 +20106,7 @@ as input for the model's response. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -20136,11 +20136,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -20150,7 +20150,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20162,11 +20162,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -20184,7 +20184,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -20283,11 +20283,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -20304,23 +20304,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -20345,7 +20345,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20353,7 +20353,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -20418,12 +20418,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -20513,7 +20513,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -20539,7 +20539,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20586,7 +20586,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -20701,7 +20701,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20709,7 +20709,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -20731,7 +20731,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20775,7 +20775,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20787,15 +20787,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -20813,7 +20813,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20849,7 +20849,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -20861,7 +20861,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -20893,7 +20893,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -20903,19 +20903,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -20929,7 +20929,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -20943,7 +20943,7 @@ as input for the model's response. - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -20992,7 +20992,7 @@ as input for the model's response. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -21023,7 +21023,7 @@ as input for the model's response. - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -21035,7 +21035,7 @@ as input for the model's response. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -21065,7 +21065,7 @@ as input for the model's response. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -21073,7 +21073,7 @@ as input for the model's response. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -21152,15 +21152,15 @@ as input for the model's response. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -21202,7 +21202,7 @@ as input for the model's response. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -21224,11 +21224,11 @@ as input for the model's response. Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -21242,11 +21242,11 @@ as input for the model's response. - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -21270,7 +21270,7 @@ as input for the model's response. - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -21278,7 +21278,7 @@ as input for the model's response. - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -21342,11 +21342,11 @@ as input for the model's response. - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -21370,11 +21370,11 @@ as input for the model's response. - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -21460,11 +21460,11 @@ as input for the model's response. - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -21510,11 +21510,11 @@ as input for the model's response. - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -21538,7 +21538,7 @@ as input for the model's response. - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -21566,11 +21566,11 @@ as input for the model's response. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -21580,7 +21580,7 @@ as input for the model's response. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -21628,11 +21628,11 @@ as input for the model's response. - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -21662,16 +21662,16 @@ as input for the model's response. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -21732,7 +21732,7 @@ as input for the model's response. The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -21782,7 +21782,7 @@ as input for the model's response. The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -21824,7 +21824,7 @@ as input for the model's response. The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -21882,7 +21882,7 @@ as input for the model's response. - `"program_output"` -- `instructions: optional string` +- `instructions: optional string or null` A system (or developer) message inserted into the model's context. @@ -21890,15 +21890,15 @@ as input for the model's response. response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses. -- `max_output_tokens: optional number` +- `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). -- `max_tool_calls: optional number` +- `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. -- `metadata: optional Metadata` +- `metadata: optional Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -22110,7 +22110,7 @@ as input for the model's response. - `"gpt-5.1-codex-max"` -- `moderation: optional object { model, policy }` +- `moderation: optional object { model, policy } or null` Configuration for running moderation on the input and output of this response. @@ -22118,11 +22118,11 @@ as input for the model's response. The moderation model to use for moderated completions, e.g. 'omni-moderation-latest'. - - `policy: optional object { input, output }` + - `policy: optional object { input, output } or null` The policy to apply to moderated response input and output. - - `input: optional object { mode }` + - `input: optional object { mode } or null` The moderation policy for the response input. @@ -22132,7 +22132,7 @@ as input for the model's response. - `"block"` - - `output: optional object { mode }` + - `output: optional object { mode } or null` The moderation policy for the response output. @@ -22142,17 +22142,17 @@ as input for the model's response. - `"block"` -- `parallel_tool_calls: optional boolean` +- `parallel_tool_calls: optional boolean or null` Whether to allow the model to run tool calls in parallel. -- `previous_response_id: optional string` +- `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. -- `prompt: optional ResponsePrompt` +- `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -22161,7 +22161,7 @@ as input for the model's response. The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -22181,11 +22181,11 @@ as input for the model's response. A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. -- `prompt_cache_key: optional string` +- `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -22207,7 +22207,7 @@ as input for the model's response. - `"30m"` -- `prompt_cache_retention: optional "in_memory" or "24h"` +- `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -22226,14 +22226,14 @@ as input for the model's response. - `"24h"` -- `reasoning: optional Reasoning` +- `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -22249,7 +22249,7 @@ as input for the model's response. - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -22273,7 +22273,7 @@ as input for the model's response. - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -22305,7 +22305,7 @@ as input for the model's response. - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -22319,12 +22319,12 @@ as input for the model's response. - `"detailed"` -- `safety_identifier: optional string` +- `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). -- `service_tier: optional "auto" or "default" or "flex" or 3 more` +- `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -22348,19 +22348,19 @@ as input for the model's response. - `"fast"` -- `store: optional boolean` +- `store: optional boolean or null` Whether to store the generated model response for later retrieval via API. -- `stream: optional boolean` +- `stream: optional boolean or null` If set to true, the model response data will be streamed to the client as it is generated using [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#Event_stream_format). See the [Streaming section below](/docs/api-reference/responses-streaming) for more information. -- `stream_options: optional object { include_obfuscation }` +- `stream_options: optional object { include_obfuscation } or null` Options for streaming responses. Only set this when you set `stream: true`. @@ -22374,7 +22374,7 @@ as input for the model's response. false to optimize for bandwidth if you trust the network links between your application and the OpenAI API. -- `temperature: optional number` +- `temperature: optional number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -22439,7 +22439,7 @@ as input for the model's response. A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -22460,7 +22460,7 @@ as input for the model's response. - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -22597,7 +22597,7 @@ as input for the model's response. - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -22671,11 +22671,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -22685,7 +22685,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -22697,11 +22697,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -22719,7 +22719,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -22818,11 +22818,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -22839,23 +22839,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -22880,7 +22880,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -22888,7 +22888,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -22953,12 +22953,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -23048,7 +23048,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -23074,7 +23074,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -23121,7 +23121,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -23236,7 +23236,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -23244,7 +23244,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -23266,7 +23266,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -23310,7 +23310,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -23322,15 +23322,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -23348,7 +23348,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -23384,7 +23384,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -23396,7 +23396,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -23428,7 +23428,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -23438,19 +23438,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -23464,7 +23464,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -23472,14 +23472,14 @@ as input for the model's response. - `"programmatic"` -- `top_logprobs: optional number` +- `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. -- `top_p: optional number` +- `top_p: optional number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -23488,7 +23488,7 @@ as input for the model's response. We generally recommend altering this or `temperature` but not both. -- `truncation: optional "auto" or "disabled"` +- `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -23520,7 +23520,7 @@ as input for the model's response. Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -23572,7 +23572,7 @@ as input for the model's response. A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -23584,7 +23584,7 @@ as input for the model's response. - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -23670,11 +23670,11 @@ as input for the model's response. - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -23712,7 +23712,7 @@ as input for the model's response. The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -23747,7 +23747,7 @@ as input for the model's response. - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -23980,7 +23980,7 @@ as input for the model's response. - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -24024,11 +24024,11 @@ as input for the model's response. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -24079,11 +24079,11 @@ as input for the model's response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -24140,7 +24140,7 @@ as input for the model's response. The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -24148,7 +24148,7 @@ as input for the model's response. A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -24195,7 +24195,7 @@ as input for the model's response. - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -24231,7 +24231,7 @@ as input for the model's response. The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -24271,7 +24271,7 @@ as input for the model's response. The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -24373,11 +24373,11 @@ as input for the model's response. - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -24385,15 +24385,15 @@ as input for the model's response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -24459,7 +24459,7 @@ as input for the model's response. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -24526,7 +24526,7 @@ as input for the model's response. The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -24595,7 +24595,7 @@ as input for the model's response. - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -24615,7 +24615,7 @@ as input for the model's response. - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -24627,15 +24627,15 @@ as input for the model's response. - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -24665,23 +24665,23 @@ as input for the model's response. - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -24697,11 +24697,11 @@ as input for the model's response. - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -24725,15 +24725,15 @@ as input for the model's response. - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -24755,11 +24755,11 @@ as input for the model's response. - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -24771,7 +24771,7 @@ as input for the model's response. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -24795,11 +24795,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -24809,7 +24809,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -24821,11 +24821,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -24843,7 +24843,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -25009,11 +25009,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -25030,23 +25030,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -25071,7 +25071,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25079,7 +25079,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -25144,12 +25144,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -25239,7 +25239,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -25297,7 +25297,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25344,7 +25344,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -25459,7 +25459,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25467,7 +25467,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -25481,7 +25481,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -25607,7 +25607,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25683,7 +25683,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25695,15 +25695,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -25721,7 +25721,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25757,7 +25757,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -25769,7 +25769,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -25801,7 +25801,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -25811,19 +25811,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -25837,7 +25837,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25851,11 +25851,11 @@ as input for the model's response. - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -25867,7 +25867,7 @@ as input for the model's response. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -25897,11 +25897,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -25911,7 +25911,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -25923,11 +25923,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -25945,7 +25945,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -26044,11 +26044,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -26065,23 +26065,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -26106,7 +26106,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -26114,7 +26114,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -26179,12 +26179,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -26274,7 +26274,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -26300,7 +26300,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -26347,7 +26347,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -26462,7 +26462,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -26470,7 +26470,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -26492,7 +26492,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -26536,7 +26536,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -26548,15 +26548,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -26574,7 +26574,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -26610,7 +26610,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -26622,7 +26622,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -26654,7 +26654,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -26664,19 +26664,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -26690,7 +26690,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -26704,7 +26704,7 @@ as input for the model's response. - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -26753,7 +26753,7 @@ as input for the model's response. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -26784,7 +26784,7 @@ as input for the model's response. - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -26796,7 +26796,7 @@ as input for the model's response. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -26826,7 +26826,7 @@ as input for the model's response. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -26834,7 +26834,7 @@ as input for the model's response. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -26913,15 +26913,15 @@ as input for the model's response. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -26963,7 +26963,7 @@ as input for the model's response. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -26985,11 +26985,11 @@ as input for the model's response. Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -27003,11 +27003,11 @@ as input for the model's response. - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -27031,7 +27031,7 @@ as input for the model's response. - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -27039,7 +27039,7 @@ as input for the model's response. - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -27103,11 +27103,11 @@ as input for the model's response. - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -27131,11 +27131,11 @@ as input for the model's response. - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -27221,11 +27221,11 @@ as input for the model's response. - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -27271,11 +27271,11 @@ as input for the model's response. - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -27299,7 +27299,7 @@ as input for the model's response. - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -27327,11 +27327,11 @@ as input for the model's response. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -27341,7 +27341,7 @@ as input for the model's response. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -27389,11 +27389,11 @@ as input for the model's response. - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -27423,16 +27423,16 @@ as input for the model's response. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -27493,7 +27493,7 @@ as input for the model's response. The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -27543,7 +27543,7 @@ as input for the model's response. The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -27585,7 +27585,7 @@ as input for the model's response. The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -27643,7 +27643,7 @@ as input for the model's response. - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -27910,11 +27910,11 @@ as input for the model's response. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -27971,7 +27971,7 @@ as input for the model's response. The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -28058,7 +28058,7 @@ as input for the model's response. - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -28150,7 +28150,7 @@ as input for the model's response. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -28211,11 +28211,11 @@ as input for the model's response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -28287,11 +28287,11 @@ as input for the model's response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -28342,7 +28342,7 @@ as input for the model's response. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -28421,7 +28421,7 @@ as input for the model's response. Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -28459,7 +28459,7 @@ as input for the model's response. The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -28493,11 +28493,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -28507,7 +28507,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -28519,11 +28519,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -28541,7 +28541,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -28640,11 +28640,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -28661,23 +28661,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -28702,7 +28702,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -28710,7 +28710,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -28775,12 +28775,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -28870,7 +28870,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -28896,7 +28896,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -28943,7 +28943,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -29058,7 +29058,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29066,7 +29066,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -29088,7 +29088,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29132,7 +29132,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29144,15 +29144,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -29170,7 +29170,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29206,7 +29206,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -29218,7 +29218,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -29250,7 +29250,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -29260,19 +29260,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -29286,7 +29286,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29342,11 +29342,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -29356,7 +29356,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29368,11 +29368,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -29390,7 +29390,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -29489,11 +29489,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -29510,23 +29510,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -29551,7 +29551,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29559,7 +29559,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -29624,12 +29624,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -29719,7 +29719,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -29745,7 +29745,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29792,7 +29792,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -29907,7 +29907,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29915,7 +29915,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -29937,7 +29937,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29981,7 +29981,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -29993,15 +29993,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -30019,7 +30019,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -30055,7 +30055,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -30067,7 +30067,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -30099,7 +30099,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -30109,19 +30109,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -30135,7 +30135,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -30179,7 +30179,7 @@ as input for the model's response. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -30209,7 +30209,7 @@ as input for the model's response. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -30217,7 +30217,7 @@ as input for the model's response. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -30296,15 +30296,15 @@ as input for the model's response. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -30346,7 +30346,7 @@ as input for the model's response. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -30370,11 +30370,11 @@ as input for the model's response. - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -30382,7 +30382,7 @@ as input for the model's response. The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -30424,7 +30424,7 @@ as input for the model's response. - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -30460,7 +30460,7 @@ as input for the model's response. The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -30524,7 +30524,7 @@ as input for the model's response. - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -30628,7 +30628,7 @@ as input for the model's response. - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -30678,7 +30678,7 @@ as input for the model's response. - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -30702,7 +30702,7 @@ as input for the model's response. The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -30732,16 +30732,16 @@ as input for the model's response. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -30783,11 +30783,11 @@ as input for the model's response. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -30797,7 +30797,7 @@ as input for the model's response. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -30849,7 +30849,7 @@ as input for the model's response. - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -30879,7 +30879,7 @@ as input for the model's response. The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -30955,7 +30955,7 @@ as input for the model's response. - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -30987,7 +30987,7 @@ as input for the model's response. Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -31116,7 +31116,7 @@ as input for the model's response. - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -31190,11 +31190,11 @@ as input for the model's response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -31204,7 +31204,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31216,11 +31216,11 @@ as input for the model's response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -31238,7 +31238,7 @@ as input for the model's response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -31337,11 +31337,11 @@ as input for the model's response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -31358,23 +31358,23 @@ as input for the model's response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -31399,7 +31399,7 @@ as input for the model's response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31407,7 +31407,7 @@ as input for the model's response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -31472,12 +31472,12 @@ as input for the model's response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -31567,7 +31567,7 @@ as input for the model's response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -31593,7 +31593,7 @@ as input for the model's response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31640,7 +31640,7 @@ as input for the model's response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -31755,7 +31755,7 @@ as input for the model's response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31763,7 +31763,7 @@ as input for the model's response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -31785,7 +31785,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31829,7 +31829,7 @@ as input for the model's response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31841,15 +31841,15 @@ as input for the model's response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -31867,7 +31867,7 @@ as input for the model's response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31903,7 +31903,7 @@ as input for the model's response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -31915,7 +31915,7 @@ as input for the model's response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -31947,7 +31947,7 @@ as input for the model's response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -31957,19 +31957,19 @@ as input for the model's response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -31983,7 +31983,7 @@ as input for the model's response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -31991,7 +31991,7 @@ as input for the model's response. - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -32000,17 +32000,17 @@ as input for the model's response. We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -32018,15 +32018,15 @@ as input for the model's response. The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -32142,19 +32142,19 @@ as input for the model's response. - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -32163,7 +32163,7 @@ as input for the model's response. The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -32183,11 +32183,11 @@ as input for the model's response. A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -32209,7 +32209,7 @@ as input for the model's response. - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -32228,14 +32228,14 @@ as input for the model's response. - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -32251,7 +32251,7 @@ as input for the model's response. - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -32275,7 +32275,7 @@ as input for the model's response. - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -32307,7 +32307,7 @@ as input for the model's response. - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -32321,12 +32321,12 @@ as input for the model's response. - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -32427,7 +32427,7 @@ as input for the model's response. A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -32448,7 +32448,7 @@ as input for the model's response. - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -32461,14 +32461,14 @@ as input for the model's response. - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -33578,7 +33578,7 @@ Retrieves a model response with the given ID. Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -33630,7 +33630,7 @@ Retrieves a model response with the given ID. A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -33642,7 +33642,7 @@ Retrieves a model response with the given ID. - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -33728,11 +33728,11 @@ Retrieves a model response with the given ID. - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -33770,7 +33770,7 @@ Retrieves a model response with the given ID. The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -33805,7 +33805,7 @@ Retrieves a model response with the given ID. - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -34038,7 +34038,7 @@ Retrieves a model response with the given ID. - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -34082,11 +34082,11 @@ Retrieves a model response with the given ID. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -34137,11 +34137,11 @@ Retrieves a model response with the given ID. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -34198,7 +34198,7 @@ Retrieves a model response with the given ID. The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -34206,7 +34206,7 @@ Retrieves a model response with the given ID. A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -34253,7 +34253,7 @@ Retrieves a model response with the given ID. - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -34289,7 +34289,7 @@ Retrieves a model response with the given ID. The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -34329,7 +34329,7 @@ Retrieves a model response with the given ID. The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -34431,11 +34431,11 @@ Retrieves a model response with the given ID. - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -34443,15 +34443,15 @@ Retrieves a model response with the given ID. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -34517,7 +34517,7 @@ Retrieves a model response with the given ID. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -34584,7 +34584,7 @@ Retrieves a model response with the given ID. The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -34653,7 +34653,7 @@ Retrieves a model response with the given ID. - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -34673,7 +34673,7 @@ Retrieves a model response with the given ID. - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -34685,15 +34685,15 @@ Retrieves a model response with the given ID. - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -34723,23 +34723,23 @@ Retrieves a model response with the given ID. - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -34755,11 +34755,11 @@ Retrieves a model response with the given ID. - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -34783,15 +34783,15 @@ Retrieves a model response with the given ID. - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -34813,11 +34813,11 @@ Retrieves a model response with the given ID. - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -34829,7 +34829,7 @@ Retrieves a model response with the given ID. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -34853,11 +34853,11 @@ Retrieves a model response with the given ID. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -34867,7 +34867,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -34879,11 +34879,11 @@ Retrieves a model response with the given ID. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -34901,7 +34901,7 @@ Retrieves a model response with the given ID. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -35067,11 +35067,11 @@ Retrieves a model response with the given ID. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -35088,23 +35088,23 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -35129,7 +35129,7 @@ Retrieves a model response with the given ID. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35137,7 +35137,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -35202,12 +35202,12 @@ Retrieves a model response with the given ID. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -35297,7 +35297,7 @@ Retrieves a model response with the given ID. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -35355,7 +35355,7 @@ Retrieves a model response with the given ID. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35402,7 +35402,7 @@ Retrieves a model response with the given ID. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -35517,7 +35517,7 @@ Retrieves a model response with the given ID. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35525,7 +35525,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -35539,7 +35539,7 @@ Retrieves a model response with the given ID. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -35665,7 +35665,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35741,7 +35741,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35753,15 +35753,15 @@ Retrieves a model response with the given ID. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -35779,7 +35779,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35815,7 +35815,7 @@ Retrieves a model response with the given ID. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -35827,7 +35827,7 @@ Retrieves a model response with the given ID. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -35859,7 +35859,7 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -35869,19 +35869,19 @@ Retrieves a model response with the given ID. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -35895,7 +35895,7 @@ Retrieves a model response with the given ID. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35909,11 +35909,11 @@ Retrieves a model response with the given ID. - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -35925,7 +35925,7 @@ Retrieves a model response with the given ID. - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -35955,11 +35955,11 @@ Retrieves a model response with the given ID. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -35969,7 +35969,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -35981,11 +35981,11 @@ Retrieves a model response with the given ID. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -36003,7 +36003,7 @@ Retrieves a model response with the given ID. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -36102,11 +36102,11 @@ Retrieves a model response with the given ID. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -36123,23 +36123,23 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -36164,7 +36164,7 @@ Retrieves a model response with the given ID. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -36172,7 +36172,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -36237,12 +36237,12 @@ Retrieves a model response with the given ID. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -36332,7 +36332,7 @@ Retrieves a model response with the given ID. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -36358,7 +36358,7 @@ Retrieves a model response with the given ID. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -36405,7 +36405,7 @@ Retrieves a model response with the given ID. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -36520,7 +36520,7 @@ Retrieves a model response with the given ID. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -36528,7 +36528,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -36550,7 +36550,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -36594,7 +36594,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -36606,15 +36606,15 @@ Retrieves a model response with the given ID. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -36632,7 +36632,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -36668,7 +36668,7 @@ Retrieves a model response with the given ID. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -36680,7 +36680,7 @@ Retrieves a model response with the given ID. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -36712,7 +36712,7 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -36722,19 +36722,19 @@ Retrieves a model response with the given ID. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -36748,7 +36748,7 @@ Retrieves a model response with the given ID. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -36762,7 +36762,7 @@ Retrieves a model response with the given ID. - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -36811,7 +36811,7 @@ Retrieves a model response with the given ID. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -36842,7 +36842,7 @@ Retrieves a model response with the given ID. - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -36854,7 +36854,7 @@ Retrieves a model response with the given ID. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -36884,7 +36884,7 @@ Retrieves a model response with the given ID. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -36892,7 +36892,7 @@ Retrieves a model response with the given ID. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -36971,15 +36971,15 @@ Retrieves a model response with the given ID. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -37021,7 +37021,7 @@ Retrieves a model response with the given ID. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -37043,11 +37043,11 @@ Retrieves a model response with the given ID. Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -37061,11 +37061,11 @@ Retrieves a model response with the given ID. - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -37089,7 +37089,7 @@ Retrieves a model response with the given ID. - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -37097,7 +37097,7 @@ Retrieves a model response with the given ID. - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -37161,11 +37161,11 @@ Retrieves a model response with the given ID. - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -37189,11 +37189,11 @@ Retrieves a model response with the given ID. - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -37279,11 +37279,11 @@ Retrieves a model response with the given ID. - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -37329,11 +37329,11 @@ Retrieves a model response with the given ID. - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -37357,7 +37357,7 @@ Retrieves a model response with the given ID. - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -37385,11 +37385,11 @@ Retrieves a model response with the given ID. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -37399,7 +37399,7 @@ Retrieves a model response with the given ID. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -37447,11 +37447,11 @@ Retrieves a model response with the given ID. - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -37481,16 +37481,16 @@ Retrieves a model response with the given ID. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -37551,7 +37551,7 @@ Retrieves a model response with the given ID. The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -37601,7 +37601,7 @@ Retrieves a model response with the given ID. The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -37643,7 +37643,7 @@ Retrieves a model response with the given ID. The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -37701,7 +37701,7 @@ Retrieves a model response with the given ID. - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -37968,11 +37968,11 @@ Retrieves a model response with the given ID. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -38029,7 +38029,7 @@ Retrieves a model response with the given ID. The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -38116,7 +38116,7 @@ Retrieves a model response with the given ID. - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -38208,7 +38208,7 @@ Retrieves a model response with the given ID. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -38269,11 +38269,11 @@ Retrieves a model response with the given ID. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -38345,11 +38345,11 @@ Retrieves a model response with the given ID. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -38400,7 +38400,7 @@ Retrieves a model response with the given ID. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -38479,7 +38479,7 @@ Retrieves a model response with the given ID. Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -38517,7 +38517,7 @@ Retrieves a model response with the given ID. The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -38551,11 +38551,11 @@ Retrieves a model response with the given ID. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -38565,7 +38565,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -38577,11 +38577,11 @@ Retrieves a model response with the given ID. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -38599,7 +38599,7 @@ Retrieves a model response with the given ID. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -38698,11 +38698,11 @@ Retrieves a model response with the given ID. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -38719,23 +38719,23 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -38760,7 +38760,7 @@ Retrieves a model response with the given ID. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -38768,7 +38768,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -38833,12 +38833,12 @@ Retrieves a model response with the given ID. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -38928,7 +38928,7 @@ Retrieves a model response with the given ID. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -38954,7 +38954,7 @@ Retrieves a model response with the given ID. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39001,7 +39001,7 @@ Retrieves a model response with the given ID. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -39116,7 +39116,7 @@ Retrieves a model response with the given ID. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39124,7 +39124,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -39146,7 +39146,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39190,7 +39190,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39202,15 +39202,15 @@ Retrieves a model response with the given ID. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -39228,7 +39228,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39264,7 +39264,7 @@ Retrieves a model response with the given ID. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -39276,7 +39276,7 @@ Retrieves a model response with the given ID. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -39308,7 +39308,7 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -39318,19 +39318,19 @@ Retrieves a model response with the given ID. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -39344,7 +39344,7 @@ Retrieves a model response with the given ID. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39400,11 +39400,11 @@ Retrieves a model response with the given ID. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -39414,7 +39414,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39426,11 +39426,11 @@ Retrieves a model response with the given ID. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -39448,7 +39448,7 @@ Retrieves a model response with the given ID. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -39547,11 +39547,11 @@ Retrieves a model response with the given ID. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -39568,23 +39568,23 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -39609,7 +39609,7 @@ Retrieves a model response with the given ID. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39617,7 +39617,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -39682,12 +39682,12 @@ Retrieves a model response with the given ID. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -39777,7 +39777,7 @@ Retrieves a model response with the given ID. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -39803,7 +39803,7 @@ Retrieves a model response with the given ID. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39850,7 +39850,7 @@ Retrieves a model response with the given ID. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -39965,7 +39965,7 @@ Retrieves a model response with the given ID. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -39973,7 +39973,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -39995,7 +39995,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -40039,7 +40039,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -40051,15 +40051,15 @@ Retrieves a model response with the given ID. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -40077,7 +40077,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -40113,7 +40113,7 @@ Retrieves a model response with the given ID. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -40125,7 +40125,7 @@ Retrieves a model response with the given ID. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -40157,7 +40157,7 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -40167,19 +40167,19 @@ Retrieves a model response with the given ID. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -40193,7 +40193,7 @@ Retrieves a model response with the given ID. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -40237,7 +40237,7 @@ Retrieves a model response with the given ID. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -40267,7 +40267,7 @@ Retrieves a model response with the given ID. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -40275,7 +40275,7 @@ Retrieves a model response with the given ID. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -40354,15 +40354,15 @@ Retrieves a model response with the given ID. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -40404,7 +40404,7 @@ Retrieves a model response with the given ID. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -40428,11 +40428,11 @@ Retrieves a model response with the given ID. - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -40440,7 +40440,7 @@ Retrieves a model response with the given ID. The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -40482,7 +40482,7 @@ Retrieves a model response with the given ID. - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -40518,7 +40518,7 @@ Retrieves a model response with the given ID. The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -40582,7 +40582,7 @@ Retrieves a model response with the given ID. - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -40686,7 +40686,7 @@ Retrieves a model response with the given ID. - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -40736,7 +40736,7 @@ Retrieves a model response with the given ID. - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -40760,7 +40760,7 @@ Retrieves a model response with the given ID. The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -40790,16 +40790,16 @@ Retrieves a model response with the given ID. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -40841,11 +40841,11 @@ Retrieves a model response with the given ID. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -40855,7 +40855,7 @@ Retrieves a model response with the given ID. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -40907,7 +40907,7 @@ Retrieves a model response with the given ID. - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -40937,7 +40937,7 @@ Retrieves a model response with the given ID. The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -41013,7 +41013,7 @@ Retrieves a model response with the given ID. - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -41045,7 +41045,7 @@ Retrieves a model response with the given ID. Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -41174,7 +41174,7 @@ Retrieves a model response with the given ID. - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -41248,11 +41248,11 @@ Retrieves a model response with the given ID. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -41262,7 +41262,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -41274,11 +41274,11 @@ Retrieves a model response with the given ID. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -41296,7 +41296,7 @@ Retrieves a model response with the given ID. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -41395,11 +41395,11 @@ Retrieves a model response with the given ID. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -41416,23 +41416,23 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -41457,7 +41457,7 @@ Retrieves a model response with the given ID. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -41465,7 +41465,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -41530,12 +41530,12 @@ Retrieves a model response with the given ID. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -41625,7 +41625,7 @@ Retrieves a model response with the given ID. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -41651,7 +41651,7 @@ Retrieves a model response with the given ID. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -41698,7 +41698,7 @@ Retrieves a model response with the given ID. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -41813,7 +41813,7 @@ Retrieves a model response with the given ID. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -41821,7 +41821,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -41843,7 +41843,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -41887,7 +41887,7 @@ Retrieves a model response with the given ID. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -41899,15 +41899,15 @@ Retrieves a model response with the given ID. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -41925,7 +41925,7 @@ Retrieves a model response with the given ID. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -41961,7 +41961,7 @@ Retrieves a model response with the given ID. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -41973,7 +41973,7 @@ Retrieves a model response with the given ID. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -42005,7 +42005,7 @@ Retrieves a model response with the given ID. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -42015,19 +42015,19 @@ Retrieves a model response with the given ID. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -42041,7 +42041,7 @@ Retrieves a model response with the given ID. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -42049,7 +42049,7 @@ Retrieves a model response with the given ID. - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -42058,17 +42058,17 @@ Retrieves a model response with the given ID. We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -42076,15 +42076,15 @@ Retrieves a model response with the given ID. The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -42200,19 +42200,19 @@ Retrieves a model response with the given ID. - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -42221,7 +42221,7 @@ Retrieves a model response with the given ID. The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -42241,11 +42241,11 @@ Retrieves a model response with the given ID. A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -42267,7 +42267,7 @@ Retrieves a model response with the given ID. - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -42286,14 +42286,14 @@ Retrieves a model response with the given ID. - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -42309,7 +42309,7 @@ Retrieves a model response with the given ID. - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -42333,7 +42333,7 @@ Retrieves a model response with the given ID. - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -42365,7 +42365,7 @@ Retrieves a model response with the given ID. - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -42379,12 +42379,12 @@ Retrieves a model response with the given ID. - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -42485,7 +42485,7 @@ Retrieves a model response with the given ID. A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -42506,7 +42506,7 @@ Retrieves a model response with the given ID. - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -42519,14 +42519,14 @@ Retrieves a model response with the given ID. - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -43098,11 +43098,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -43132,11 +43132,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: string` + - `file_id: string or null` The identifier of an uploaded file that contains the screenshot. - - `image_url: string` + - `image_url: string or null` The URL of the screenshot image. @@ -43180,7 +43180,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -43238,7 +43238,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend phase on all assistant messages — dropping it can degrade performance. Not used for user messages. @@ -43325,7 +43325,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -43370,7 +43370,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -43408,7 +43408,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -43442,11 +43442,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -43456,7 +43456,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -43468,11 +43468,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -43490,7 +43490,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -43656,11 +43656,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -43677,23 +43677,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -43718,7 +43718,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -43726,7 +43726,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -43791,12 +43791,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -43886,7 +43886,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -43944,7 +43944,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -43991,7 +43991,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -44106,7 +44106,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44114,7 +44114,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -44128,7 +44128,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -44254,7 +44254,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44330,7 +44330,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44342,15 +44342,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -44368,7 +44368,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44404,7 +44404,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -44416,7 +44416,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -44448,7 +44448,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -44458,19 +44458,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -44484,7 +44484,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44540,11 +44540,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -44554,7 +44554,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44566,11 +44566,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -44588,7 +44588,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -44687,11 +44687,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -44708,23 +44708,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -44749,7 +44749,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44757,7 +44757,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -44822,12 +44822,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -44917,7 +44917,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -44943,7 +44943,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -44990,7 +44990,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -45105,7 +45105,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -45113,7 +45113,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -45135,7 +45135,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -45179,7 +45179,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -45191,15 +45191,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -45217,7 +45217,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -45253,7 +45253,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -45265,7 +45265,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -45297,7 +45297,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -45307,19 +45307,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -45333,7 +45333,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -45391,7 +45391,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -45468,11 +45468,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -45558,7 +45558,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -45606,7 +45606,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -45649,11 +45649,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -45710,7 +45710,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -45718,7 +45718,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -45765,7 +45765,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -45801,7 +45801,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -45841,7 +45841,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -45967,11 +45967,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -46022,7 +46022,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -46069,7 +46069,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -46077,7 +46077,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -46156,15 +46156,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -46206,7 +46206,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -46230,11 +46230,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -46242,7 +46242,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -46284,7 +46284,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -46320,7 +46320,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -46384,7 +46384,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -46488,7 +46488,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -46538,7 +46538,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -46562,7 +46562,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -46590,11 +46590,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -46604,7 +46604,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -46656,7 +46656,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -46686,16 +46686,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -46739,7 +46739,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -46806,7 +46806,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -46905,7 +46905,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -46913,7 +46913,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -46960,7 +46960,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -46996,7 +46996,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -47036,7 +47036,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -47103,7 +47103,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -47111,7 +47111,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -47158,7 +47158,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -47194,7 +47194,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -47234,7 +47234,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -47276,7 +47276,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -47524,11 +47524,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -47566,7 +47566,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -47601,7 +47601,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -47731,7 +47731,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -47783,7 +47783,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -47795,7 +47795,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -47881,11 +47881,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -47923,7 +47923,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -47958,7 +47958,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -48191,7 +48191,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -48235,11 +48235,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -48290,11 +48290,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -48351,7 +48351,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -48359,7 +48359,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -48406,7 +48406,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -48442,7 +48442,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -48482,7 +48482,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -48584,11 +48584,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -48596,15 +48596,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -48670,7 +48670,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -48737,7 +48737,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -48806,7 +48806,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -48826,7 +48826,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -48838,15 +48838,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -48876,23 +48876,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -48908,11 +48908,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -48936,15 +48936,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -48966,11 +48966,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -48982,7 +48982,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -49006,11 +49006,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -49020,7 +49020,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -49032,11 +49032,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -49054,7 +49054,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -49220,11 +49220,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -49241,23 +49241,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -49282,7 +49282,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -49290,7 +49290,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -49355,12 +49355,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -49450,7 +49450,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -49508,7 +49508,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -49555,7 +49555,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -49670,7 +49670,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -49678,7 +49678,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -49692,7 +49692,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -49818,7 +49818,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -49894,7 +49894,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -49906,15 +49906,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -49932,7 +49932,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -49968,7 +49968,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -49980,7 +49980,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -50012,7 +50012,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -50022,19 +50022,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -50048,7 +50048,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50062,11 +50062,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -50078,7 +50078,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -50108,11 +50108,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -50122,7 +50122,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50134,11 +50134,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -50156,7 +50156,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -50255,11 +50255,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -50276,23 +50276,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -50317,7 +50317,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50325,7 +50325,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -50390,12 +50390,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -50485,7 +50485,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -50511,7 +50511,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50558,7 +50558,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -50673,7 +50673,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50681,7 +50681,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -50703,7 +50703,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50747,7 +50747,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50759,15 +50759,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -50785,7 +50785,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50821,7 +50821,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -50833,7 +50833,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -50865,7 +50865,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -50875,19 +50875,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -50901,7 +50901,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -50915,7 +50915,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -50964,7 +50964,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -50995,7 +50995,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -51007,7 +51007,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -51037,7 +51037,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -51045,7 +51045,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -51124,15 +51124,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -51174,7 +51174,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -51196,11 +51196,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -51214,11 +51214,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -51242,7 +51242,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -51250,7 +51250,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -51314,11 +51314,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -51342,11 +51342,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -51432,11 +51432,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -51482,11 +51482,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -51510,7 +51510,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -51538,11 +51538,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -51552,7 +51552,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -51600,11 +51600,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -51634,16 +51634,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -51704,7 +51704,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -51754,7 +51754,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -51796,7 +51796,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -51854,7 +51854,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -52121,11 +52121,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -52182,7 +52182,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -52269,7 +52269,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -52361,7 +52361,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -52422,11 +52422,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -52498,11 +52498,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -52553,7 +52553,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -52632,7 +52632,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -52670,7 +52670,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -52704,11 +52704,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -52718,7 +52718,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -52730,11 +52730,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -52752,7 +52752,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -52851,11 +52851,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -52872,23 +52872,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -52913,7 +52913,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -52921,7 +52921,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -52986,12 +52986,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -53081,7 +53081,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -53107,7 +53107,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53154,7 +53154,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -53269,7 +53269,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53277,7 +53277,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -53299,7 +53299,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53343,7 +53343,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53355,15 +53355,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -53381,7 +53381,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53417,7 +53417,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -53429,7 +53429,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -53461,7 +53461,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -53471,19 +53471,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -53497,7 +53497,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53553,11 +53553,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -53567,7 +53567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53579,11 +53579,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -53601,7 +53601,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -53700,11 +53700,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -53721,23 +53721,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -53762,7 +53762,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -53770,7 +53770,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -53835,12 +53835,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -53930,7 +53930,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -53956,7 +53956,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -54003,7 +54003,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -54118,7 +54118,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -54126,7 +54126,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -54148,7 +54148,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -54192,7 +54192,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -54204,15 +54204,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -54230,7 +54230,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -54266,7 +54266,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -54278,7 +54278,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -54310,7 +54310,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -54320,19 +54320,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -54346,7 +54346,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -54390,7 +54390,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -54420,7 +54420,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -54428,7 +54428,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -54507,15 +54507,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -54557,7 +54557,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -54581,11 +54581,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -54593,7 +54593,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -54635,7 +54635,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -54671,7 +54671,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -54735,7 +54735,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -54839,7 +54839,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -54889,7 +54889,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -54913,7 +54913,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -54943,16 +54943,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -54994,11 +54994,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -55008,7 +55008,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -55060,7 +55060,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -55090,7 +55090,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -55166,7 +55166,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -55198,7 +55198,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -55327,7 +55327,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -55401,11 +55401,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -55415,7 +55415,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -55427,11 +55427,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -55449,7 +55449,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -55548,11 +55548,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -55569,23 +55569,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -55610,7 +55610,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -55618,7 +55618,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -55683,12 +55683,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -55778,7 +55778,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -55804,7 +55804,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -55851,7 +55851,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -55966,7 +55966,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -55974,7 +55974,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -55996,7 +55996,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -56040,7 +56040,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -56052,15 +56052,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -56078,7 +56078,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -56114,7 +56114,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -56126,7 +56126,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -56158,7 +56158,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -56168,19 +56168,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -56194,7 +56194,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -56202,7 +56202,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -56211,17 +56211,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -56229,15 +56229,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -56353,19 +56353,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -56374,7 +56374,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -56394,11 +56394,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -56420,7 +56420,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -56439,14 +56439,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -56462,7 +56462,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -56486,7 +56486,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -56518,7 +56518,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -56532,12 +56532,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -56638,7 +56638,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -56659,7 +56659,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -56672,14 +56672,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -56955,7 +56955,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -57007,7 +57007,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -57019,7 +57019,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -57105,11 +57105,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -57147,7 +57147,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -57182,7 +57182,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -57415,7 +57415,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -57459,11 +57459,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -57514,11 +57514,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -57575,7 +57575,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -57583,7 +57583,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -57630,7 +57630,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -57666,7 +57666,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -57706,7 +57706,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -57808,11 +57808,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -57820,15 +57820,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -57894,7 +57894,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -57961,7 +57961,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -58030,7 +58030,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -58050,7 +58050,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -58062,15 +58062,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -58100,23 +58100,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -58132,11 +58132,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -58160,15 +58160,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -58190,11 +58190,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -58206,7 +58206,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -58230,11 +58230,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -58244,7 +58244,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -58256,11 +58256,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -58278,7 +58278,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -58444,11 +58444,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -58465,23 +58465,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -58506,7 +58506,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -58514,7 +58514,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -58579,12 +58579,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -58674,7 +58674,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -58732,7 +58732,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -58779,7 +58779,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -58894,7 +58894,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -58902,7 +58902,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -58916,7 +58916,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -59042,7 +59042,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59118,7 +59118,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59130,15 +59130,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -59156,7 +59156,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59192,7 +59192,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -59204,7 +59204,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -59236,7 +59236,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -59246,19 +59246,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -59272,7 +59272,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59286,11 +59286,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -59302,7 +59302,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -59332,11 +59332,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -59346,7 +59346,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59358,11 +59358,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -59380,7 +59380,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -59479,11 +59479,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -59500,23 +59500,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -59541,7 +59541,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59549,7 +59549,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -59614,12 +59614,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -59709,7 +59709,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -59735,7 +59735,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59782,7 +59782,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -59897,7 +59897,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59905,7 +59905,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -59927,7 +59927,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59971,7 +59971,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -59983,15 +59983,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -60009,7 +60009,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -60045,7 +60045,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -60057,7 +60057,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -60089,7 +60089,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -60099,19 +60099,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -60125,7 +60125,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -60139,7 +60139,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -60188,7 +60188,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -60219,7 +60219,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -60231,7 +60231,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -60261,7 +60261,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -60269,7 +60269,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -60348,15 +60348,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -60398,7 +60398,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -60420,11 +60420,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -60438,11 +60438,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -60466,7 +60466,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -60474,7 +60474,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -60538,11 +60538,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -60566,11 +60566,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -60656,11 +60656,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -60706,11 +60706,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -60734,7 +60734,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -60762,11 +60762,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -60776,7 +60776,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -60824,11 +60824,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -60858,16 +60858,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -60928,7 +60928,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -60978,7 +60978,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -61020,7 +61020,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -61078,7 +61078,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -61345,11 +61345,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -61406,7 +61406,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -61493,7 +61493,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -61585,7 +61585,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -61646,11 +61646,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -61722,11 +61722,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -61777,7 +61777,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -61856,7 +61856,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -61894,7 +61894,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -61928,11 +61928,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -61942,7 +61942,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -61954,11 +61954,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -61976,7 +61976,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -62075,11 +62075,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -62096,23 +62096,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -62137,7 +62137,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62145,7 +62145,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -62210,12 +62210,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -62305,7 +62305,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -62331,7 +62331,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62378,7 +62378,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -62493,7 +62493,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62501,7 +62501,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -62523,7 +62523,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62567,7 +62567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62579,15 +62579,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -62605,7 +62605,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62641,7 +62641,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -62653,7 +62653,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -62685,7 +62685,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -62695,19 +62695,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -62721,7 +62721,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62777,11 +62777,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -62791,7 +62791,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62803,11 +62803,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -62825,7 +62825,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -62924,11 +62924,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -62945,23 +62945,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -62986,7 +62986,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -62994,7 +62994,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -63059,12 +63059,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -63154,7 +63154,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -63180,7 +63180,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -63227,7 +63227,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -63342,7 +63342,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -63350,7 +63350,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -63372,7 +63372,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -63416,7 +63416,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -63428,15 +63428,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -63454,7 +63454,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -63490,7 +63490,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -63502,7 +63502,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -63534,7 +63534,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -63544,19 +63544,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -63570,7 +63570,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -63614,7 +63614,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -63644,7 +63644,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -63652,7 +63652,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -63731,15 +63731,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -63781,7 +63781,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -63805,11 +63805,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -63817,7 +63817,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -63859,7 +63859,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -63895,7 +63895,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -63959,7 +63959,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -64063,7 +64063,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -64113,7 +64113,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -64137,7 +64137,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -64167,16 +64167,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -64218,11 +64218,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -64232,7 +64232,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -64284,7 +64284,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -64314,7 +64314,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -64390,7 +64390,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -64422,7 +64422,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -64551,7 +64551,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -64625,11 +64625,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -64639,7 +64639,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -64651,11 +64651,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -64673,7 +64673,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -64772,11 +64772,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -64793,23 +64793,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -64834,7 +64834,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -64842,7 +64842,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -64907,12 +64907,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -65002,7 +65002,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -65028,7 +65028,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -65075,7 +65075,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -65190,7 +65190,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -65198,7 +65198,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -65220,7 +65220,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -65264,7 +65264,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -65276,15 +65276,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -65302,7 +65302,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -65338,7 +65338,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -65350,7 +65350,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -65382,7 +65382,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -65392,19 +65392,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -65418,7 +65418,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -65426,7 +65426,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -65435,17 +65435,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -65453,15 +65453,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -65577,19 +65577,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -65598,7 +65598,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -65618,11 +65618,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -65644,7 +65644,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -65663,14 +65663,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -65686,7 +65686,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -65710,7 +65710,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -65742,7 +65742,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -65756,12 +65756,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -65862,7 +65862,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -65883,7 +65883,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -65896,14 +65896,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -66058,11 +66058,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -66100,7 +66100,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -66688,7 +66688,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -66740,7 +66740,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -66752,7 +66752,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -66838,11 +66838,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -66880,7 +66880,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -66915,7 +66915,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -67148,7 +67148,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -67192,11 +67192,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -67247,11 +67247,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -67308,7 +67308,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -67316,7 +67316,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -67363,7 +67363,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -67399,7 +67399,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -67439,7 +67439,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -67541,11 +67541,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -67553,15 +67553,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -67627,7 +67627,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -67694,7 +67694,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -67763,7 +67763,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -67783,7 +67783,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -67795,15 +67795,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -67833,23 +67833,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -67865,11 +67865,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -67893,15 +67893,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -67923,11 +67923,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -67939,7 +67939,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -67963,11 +67963,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -67977,7 +67977,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -67989,11 +67989,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -68011,7 +68011,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -68177,11 +68177,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -68198,23 +68198,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -68239,7 +68239,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -68247,7 +68247,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -68312,12 +68312,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -68407,7 +68407,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -68465,7 +68465,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -68512,7 +68512,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -68627,7 +68627,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -68635,7 +68635,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -68649,7 +68649,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -68775,7 +68775,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -68851,7 +68851,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -68863,15 +68863,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -68889,7 +68889,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -68925,7 +68925,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -68937,7 +68937,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -68969,7 +68969,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -68979,19 +68979,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -69005,7 +69005,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69019,11 +69019,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -69035,7 +69035,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -69065,11 +69065,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -69079,7 +69079,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69091,11 +69091,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -69113,7 +69113,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -69212,11 +69212,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -69233,23 +69233,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -69274,7 +69274,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69282,7 +69282,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -69347,12 +69347,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -69442,7 +69442,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -69468,7 +69468,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69515,7 +69515,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -69630,7 +69630,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69638,7 +69638,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -69660,7 +69660,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69704,7 +69704,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69716,15 +69716,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -69742,7 +69742,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69778,7 +69778,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -69790,7 +69790,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -69822,7 +69822,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -69832,19 +69832,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -69858,7 +69858,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -69872,7 +69872,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -69921,7 +69921,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -69952,7 +69952,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -69964,7 +69964,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -69994,7 +69994,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -70002,7 +70002,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -70081,15 +70081,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -70131,7 +70131,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -70153,11 +70153,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -70171,11 +70171,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -70199,7 +70199,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -70207,7 +70207,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -70271,11 +70271,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -70299,11 +70299,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -70389,11 +70389,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -70439,11 +70439,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -70467,7 +70467,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -70495,11 +70495,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -70509,7 +70509,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -70557,11 +70557,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -70591,16 +70591,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -70661,7 +70661,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -70711,7 +70711,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -70753,7 +70753,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -70811,7 +70811,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -71078,11 +71078,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -71139,7 +71139,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -71226,7 +71226,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -71318,7 +71318,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -71379,11 +71379,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -71455,11 +71455,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -71510,7 +71510,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -71589,7 +71589,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -71627,7 +71627,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -71661,11 +71661,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -71675,7 +71675,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -71687,11 +71687,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -71709,7 +71709,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -71808,11 +71808,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -71829,23 +71829,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -71870,7 +71870,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -71878,7 +71878,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -71943,12 +71943,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -72038,7 +72038,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -72064,7 +72064,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72111,7 +72111,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -72226,7 +72226,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72234,7 +72234,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -72256,7 +72256,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72300,7 +72300,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72312,15 +72312,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -72338,7 +72338,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72374,7 +72374,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -72386,7 +72386,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -72418,7 +72418,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -72428,19 +72428,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -72454,7 +72454,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72510,11 +72510,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -72524,7 +72524,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72536,11 +72536,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -72558,7 +72558,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -72657,11 +72657,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -72678,23 +72678,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -72719,7 +72719,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72727,7 +72727,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -72792,12 +72792,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -72887,7 +72887,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -72913,7 +72913,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -72960,7 +72960,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -73075,7 +73075,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -73083,7 +73083,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -73105,7 +73105,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -73149,7 +73149,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -73161,15 +73161,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -73187,7 +73187,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -73223,7 +73223,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -73235,7 +73235,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -73267,7 +73267,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -73277,19 +73277,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -73303,7 +73303,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -73347,7 +73347,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -73377,7 +73377,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -73385,7 +73385,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -73464,15 +73464,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -73514,7 +73514,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -73538,11 +73538,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -73550,7 +73550,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -73592,7 +73592,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -73628,7 +73628,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -73692,7 +73692,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -73796,7 +73796,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -73846,7 +73846,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -73870,7 +73870,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -73900,16 +73900,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -73951,11 +73951,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -73965,7 +73965,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -74017,7 +74017,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -74047,7 +74047,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -74123,7 +74123,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -74155,7 +74155,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -74284,7 +74284,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -74358,11 +74358,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -74372,7 +74372,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -74384,11 +74384,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -74406,7 +74406,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -74505,11 +74505,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -74526,23 +74526,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -74567,7 +74567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -74575,7 +74575,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -74640,12 +74640,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -74735,7 +74735,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -74761,7 +74761,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -74808,7 +74808,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -74923,7 +74923,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -74931,7 +74931,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -74953,7 +74953,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -74997,7 +74997,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -75009,15 +75009,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -75035,7 +75035,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -75071,7 +75071,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -75083,7 +75083,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -75115,7 +75115,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -75125,19 +75125,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -75151,7 +75151,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -75159,7 +75159,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -75168,17 +75168,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -75186,15 +75186,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -75310,19 +75310,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -75331,7 +75331,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -75351,11 +75351,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -75377,7 +75377,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -75396,14 +75396,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -75419,7 +75419,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -75443,7 +75443,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -75475,7 +75475,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -75489,12 +75489,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -75595,7 +75595,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -75616,7 +75616,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -75629,14 +75629,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -75820,7 +75820,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Emitted when an error occurs. - - `code: string` + - `code: string or null` The error code. @@ -75828,7 +75828,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The error message. - - `param: string` + - `param: string or null` The error parameter. @@ -75860,7 +75860,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -75912,7 +75912,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -75924,7 +75924,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -76010,11 +76010,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -76052,7 +76052,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -76087,7 +76087,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -76320,7 +76320,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -76364,11 +76364,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -76419,11 +76419,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -76480,7 +76480,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -76488,7 +76488,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -76535,7 +76535,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -76571,7 +76571,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -76611,7 +76611,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -76713,11 +76713,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -76725,15 +76725,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -76799,7 +76799,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -76866,7 +76866,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -76935,7 +76935,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -76955,7 +76955,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -76967,15 +76967,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -77005,23 +77005,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -77037,11 +77037,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -77065,15 +77065,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -77095,11 +77095,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -77111,7 +77111,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -77135,11 +77135,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -77149,7 +77149,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -77161,11 +77161,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -77183,7 +77183,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -77349,11 +77349,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -77370,23 +77370,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -77411,7 +77411,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -77419,7 +77419,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -77484,12 +77484,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -77579,7 +77579,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -77637,7 +77637,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -77684,7 +77684,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -77799,7 +77799,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -77807,7 +77807,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -77821,7 +77821,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -77947,7 +77947,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78023,7 +78023,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78035,15 +78035,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -78061,7 +78061,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78097,7 +78097,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -78109,7 +78109,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -78141,7 +78141,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -78151,19 +78151,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -78177,7 +78177,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78191,11 +78191,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -78207,7 +78207,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -78237,11 +78237,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -78251,7 +78251,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78263,11 +78263,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -78285,7 +78285,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -78384,11 +78384,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -78405,23 +78405,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -78446,7 +78446,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78454,7 +78454,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -78519,12 +78519,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -78614,7 +78614,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -78640,7 +78640,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78687,7 +78687,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -78802,7 +78802,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78810,7 +78810,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -78832,7 +78832,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78876,7 +78876,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78888,15 +78888,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -78914,7 +78914,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -78950,7 +78950,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -78962,7 +78962,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -78994,7 +78994,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -79004,19 +79004,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -79030,7 +79030,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -79044,7 +79044,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -79093,7 +79093,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -79124,7 +79124,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -79136,7 +79136,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -79166,7 +79166,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -79174,7 +79174,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -79253,15 +79253,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -79303,7 +79303,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -79325,11 +79325,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -79343,11 +79343,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -79371,7 +79371,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -79379,7 +79379,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -79443,11 +79443,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -79471,11 +79471,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -79561,11 +79561,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -79611,11 +79611,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -79639,7 +79639,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -79667,11 +79667,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -79681,7 +79681,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -79729,11 +79729,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -79763,16 +79763,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -79833,7 +79833,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -79883,7 +79883,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -79925,7 +79925,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -79983,7 +79983,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -80250,11 +80250,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -80311,7 +80311,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -80398,7 +80398,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -80490,7 +80490,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -80551,11 +80551,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -80627,11 +80627,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -80682,7 +80682,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -80761,7 +80761,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -80799,7 +80799,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -80833,11 +80833,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -80847,7 +80847,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -80859,11 +80859,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -80881,7 +80881,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -80980,11 +80980,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -81001,23 +81001,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -81042,7 +81042,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81050,7 +81050,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -81115,12 +81115,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -81210,7 +81210,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -81236,7 +81236,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81283,7 +81283,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -81398,7 +81398,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81406,7 +81406,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -81428,7 +81428,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81472,7 +81472,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81484,15 +81484,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -81510,7 +81510,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81546,7 +81546,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -81558,7 +81558,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -81590,7 +81590,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -81600,19 +81600,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -81626,7 +81626,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81682,11 +81682,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -81696,7 +81696,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81708,11 +81708,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -81730,7 +81730,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -81829,11 +81829,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -81850,23 +81850,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -81891,7 +81891,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -81899,7 +81899,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -81964,12 +81964,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -82059,7 +82059,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -82085,7 +82085,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -82132,7 +82132,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -82247,7 +82247,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -82255,7 +82255,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -82277,7 +82277,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -82321,7 +82321,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -82333,15 +82333,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -82359,7 +82359,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -82395,7 +82395,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -82407,7 +82407,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -82439,7 +82439,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -82449,19 +82449,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -82475,7 +82475,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -82519,7 +82519,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -82549,7 +82549,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -82557,7 +82557,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -82636,15 +82636,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -82686,7 +82686,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -82710,11 +82710,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -82722,7 +82722,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -82764,7 +82764,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -82800,7 +82800,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -82864,7 +82864,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -82968,7 +82968,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -83018,7 +83018,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -83042,7 +83042,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -83072,16 +83072,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -83123,11 +83123,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -83137,7 +83137,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -83189,7 +83189,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -83219,7 +83219,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -83295,7 +83295,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -83327,7 +83327,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -83456,7 +83456,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -83530,11 +83530,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -83544,7 +83544,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -83556,11 +83556,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -83578,7 +83578,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -83677,11 +83677,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -83698,23 +83698,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -83739,7 +83739,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -83747,7 +83747,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -83812,12 +83812,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -83907,7 +83907,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -83933,7 +83933,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -83980,7 +83980,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -84095,7 +84095,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -84103,7 +84103,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -84125,7 +84125,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -84169,7 +84169,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -84181,15 +84181,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -84207,7 +84207,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -84243,7 +84243,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -84255,7 +84255,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -84287,7 +84287,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -84297,19 +84297,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -84323,7 +84323,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -84331,7 +84331,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -84340,17 +84340,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -84358,15 +84358,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -84482,19 +84482,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -84503,7 +84503,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -84523,11 +84523,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -84549,7 +84549,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -84568,14 +84568,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -84591,7 +84591,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -84615,7 +84615,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -84647,7 +84647,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -84661,12 +84661,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -84767,7 +84767,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -84788,7 +84788,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -84801,14 +84801,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -85002,7 +85002,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -85051,7 +85051,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -85281,7 +85281,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -85333,7 +85333,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -85345,7 +85345,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -85431,11 +85431,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -85473,7 +85473,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -85508,7 +85508,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -85741,7 +85741,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -85785,11 +85785,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -85840,11 +85840,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -85901,7 +85901,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -85909,7 +85909,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -85956,7 +85956,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -85992,7 +85992,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -86032,7 +86032,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -86134,11 +86134,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -86146,15 +86146,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -86220,7 +86220,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -86287,7 +86287,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -86356,7 +86356,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -86376,7 +86376,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -86388,15 +86388,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -86426,23 +86426,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -86458,11 +86458,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -86486,15 +86486,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -86516,11 +86516,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -86532,7 +86532,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -86556,11 +86556,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -86570,7 +86570,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -86582,11 +86582,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -86604,7 +86604,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -86770,11 +86770,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -86791,23 +86791,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -86832,7 +86832,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -86840,7 +86840,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -86905,12 +86905,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -87000,7 +87000,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -87058,7 +87058,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87105,7 +87105,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -87220,7 +87220,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87228,7 +87228,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -87242,7 +87242,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -87368,7 +87368,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87444,7 +87444,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87456,15 +87456,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -87482,7 +87482,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87518,7 +87518,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -87530,7 +87530,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -87562,7 +87562,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -87572,19 +87572,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -87598,7 +87598,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87612,11 +87612,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -87628,7 +87628,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -87658,11 +87658,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -87672,7 +87672,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87684,11 +87684,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -87706,7 +87706,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -87805,11 +87805,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -87826,23 +87826,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -87867,7 +87867,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -87875,7 +87875,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -87940,12 +87940,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -88035,7 +88035,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -88061,7 +88061,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -88108,7 +88108,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -88223,7 +88223,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -88231,7 +88231,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -88253,7 +88253,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -88297,7 +88297,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -88309,15 +88309,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -88335,7 +88335,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -88371,7 +88371,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -88383,7 +88383,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -88415,7 +88415,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -88425,19 +88425,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -88451,7 +88451,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -88465,7 +88465,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -88514,7 +88514,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -88545,7 +88545,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -88557,7 +88557,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -88587,7 +88587,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -88595,7 +88595,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -88674,15 +88674,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -88724,7 +88724,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -88746,11 +88746,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -88764,11 +88764,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -88792,7 +88792,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -88800,7 +88800,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -88864,11 +88864,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -88892,11 +88892,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -88982,11 +88982,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -89032,11 +89032,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -89060,7 +89060,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -89088,11 +89088,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -89102,7 +89102,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -89150,11 +89150,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -89184,16 +89184,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -89254,7 +89254,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -89304,7 +89304,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -89346,7 +89346,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -89404,7 +89404,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -89671,11 +89671,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -89732,7 +89732,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -89819,7 +89819,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -89911,7 +89911,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -89972,11 +89972,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -90048,11 +90048,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -90103,7 +90103,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -90182,7 +90182,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -90220,7 +90220,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -90254,11 +90254,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -90268,7 +90268,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -90280,11 +90280,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -90302,7 +90302,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -90401,11 +90401,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -90422,23 +90422,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -90463,7 +90463,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -90471,7 +90471,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -90536,12 +90536,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -90631,7 +90631,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -90657,7 +90657,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -90704,7 +90704,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -90819,7 +90819,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -90827,7 +90827,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -90849,7 +90849,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -90893,7 +90893,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -90905,15 +90905,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -90931,7 +90931,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -90967,7 +90967,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -90979,7 +90979,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -91011,7 +91011,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -91021,19 +91021,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -91047,7 +91047,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91103,11 +91103,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -91117,7 +91117,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91129,11 +91129,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -91151,7 +91151,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -91250,11 +91250,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -91271,23 +91271,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -91312,7 +91312,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91320,7 +91320,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -91385,12 +91385,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -91480,7 +91480,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -91506,7 +91506,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91553,7 +91553,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -91668,7 +91668,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91676,7 +91676,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -91698,7 +91698,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91742,7 +91742,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91754,15 +91754,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -91780,7 +91780,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91816,7 +91816,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -91828,7 +91828,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -91860,7 +91860,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -91870,19 +91870,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -91896,7 +91896,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -91940,7 +91940,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -91970,7 +91970,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -91978,7 +91978,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -92057,15 +92057,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -92107,7 +92107,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -92131,11 +92131,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -92143,7 +92143,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -92185,7 +92185,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -92221,7 +92221,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -92285,7 +92285,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -92389,7 +92389,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -92439,7 +92439,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -92463,7 +92463,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -92493,16 +92493,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -92544,11 +92544,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -92558,7 +92558,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -92610,7 +92610,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -92640,7 +92640,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -92716,7 +92716,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -92748,7 +92748,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -92877,7 +92877,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -92951,11 +92951,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -92965,7 +92965,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -92977,11 +92977,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -92999,7 +92999,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -93098,11 +93098,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -93119,23 +93119,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -93160,7 +93160,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -93168,7 +93168,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -93233,12 +93233,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -93328,7 +93328,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -93354,7 +93354,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -93401,7 +93401,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -93516,7 +93516,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -93524,7 +93524,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -93546,7 +93546,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -93590,7 +93590,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -93602,15 +93602,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -93628,7 +93628,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -93664,7 +93664,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -93676,7 +93676,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -93708,7 +93708,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -93718,19 +93718,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -93744,7 +93744,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -93752,7 +93752,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -93761,17 +93761,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -93779,15 +93779,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -93903,19 +93903,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -93924,7 +93924,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -93944,11 +93944,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -93970,7 +93970,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -93989,14 +93989,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -94012,7 +94012,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -94036,7 +94036,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -94068,7 +94068,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -94082,12 +94082,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -94188,7 +94188,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -94209,7 +94209,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -94222,14 +94222,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -94346,7 +94346,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -94398,7 +94398,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -94410,7 +94410,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -94496,11 +94496,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -94538,7 +94538,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -94573,7 +94573,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -94806,7 +94806,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -94850,11 +94850,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -94905,11 +94905,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -94966,7 +94966,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -94974,7 +94974,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -95021,7 +95021,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -95057,7 +95057,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -95097,7 +95097,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -95199,11 +95199,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -95211,15 +95211,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -95285,7 +95285,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -95352,7 +95352,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -95421,7 +95421,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -95441,7 +95441,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -95453,15 +95453,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -95491,23 +95491,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -95523,11 +95523,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -95551,15 +95551,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -95581,11 +95581,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -95597,7 +95597,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -95621,11 +95621,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -95635,7 +95635,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -95647,11 +95647,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -95669,7 +95669,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -95835,11 +95835,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -95856,23 +95856,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -95897,7 +95897,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -95905,7 +95905,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -95970,12 +95970,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -96065,7 +96065,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -96123,7 +96123,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96170,7 +96170,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -96285,7 +96285,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96293,7 +96293,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -96307,7 +96307,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -96433,7 +96433,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96509,7 +96509,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96521,15 +96521,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -96547,7 +96547,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96583,7 +96583,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -96595,7 +96595,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -96627,7 +96627,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -96637,19 +96637,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -96663,7 +96663,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96677,11 +96677,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -96693,7 +96693,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -96723,11 +96723,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -96737,7 +96737,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96749,11 +96749,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -96771,7 +96771,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -96870,11 +96870,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -96891,23 +96891,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -96932,7 +96932,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -96940,7 +96940,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -97005,12 +97005,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -97100,7 +97100,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -97126,7 +97126,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -97173,7 +97173,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -97288,7 +97288,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -97296,7 +97296,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -97318,7 +97318,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -97362,7 +97362,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -97374,15 +97374,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -97400,7 +97400,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -97436,7 +97436,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -97448,7 +97448,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -97480,7 +97480,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -97490,19 +97490,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -97516,7 +97516,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -97530,7 +97530,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -97579,7 +97579,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -97610,7 +97610,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -97622,7 +97622,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -97652,7 +97652,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -97660,7 +97660,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -97739,15 +97739,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -97789,7 +97789,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -97811,11 +97811,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -97829,11 +97829,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -97857,7 +97857,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -97865,7 +97865,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -97929,11 +97929,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -97957,11 +97957,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -98047,11 +98047,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -98097,11 +98097,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -98125,7 +98125,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -98153,11 +98153,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -98167,7 +98167,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -98215,11 +98215,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -98249,16 +98249,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -98319,7 +98319,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -98369,7 +98369,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -98411,7 +98411,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -98469,7 +98469,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -98736,11 +98736,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -98797,7 +98797,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -98884,7 +98884,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -98976,7 +98976,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -99037,11 +99037,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -99113,11 +99113,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -99168,7 +99168,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -99247,7 +99247,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -99285,7 +99285,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -99319,11 +99319,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -99333,7 +99333,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -99345,11 +99345,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -99367,7 +99367,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -99466,11 +99466,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -99487,23 +99487,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -99528,7 +99528,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -99536,7 +99536,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -99601,12 +99601,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -99696,7 +99696,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -99722,7 +99722,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -99769,7 +99769,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -99884,7 +99884,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -99892,7 +99892,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -99914,7 +99914,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -99958,7 +99958,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -99970,15 +99970,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -99996,7 +99996,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100032,7 +100032,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -100044,7 +100044,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -100076,7 +100076,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -100086,19 +100086,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -100112,7 +100112,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100168,11 +100168,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -100182,7 +100182,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100194,11 +100194,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -100216,7 +100216,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -100315,11 +100315,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -100336,23 +100336,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -100377,7 +100377,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100385,7 +100385,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -100450,12 +100450,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -100545,7 +100545,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -100571,7 +100571,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100618,7 +100618,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -100733,7 +100733,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100741,7 +100741,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -100763,7 +100763,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100807,7 +100807,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100819,15 +100819,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -100845,7 +100845,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -100881,7 +100881,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -100893,7 +100893,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -100925,7 +100925,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -100935,19 +100935,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -100961,7 +100961,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -101005,7 +101005,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -101035,7 +101035,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -101043,7 +101043,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -101122,15 +101122,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -101172,7 +101172,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -101196,11 +101196,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -101208,7 +101208,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -101250,7 +101250,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -101286,7 +101286,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -101350,7 +101350,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -101454,7 +101454,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -101504,7 +101504,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -101528,7 +101528,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -101558,16 +101558,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -101609,11 +101609,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -101623,7 +101623,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -101675,7 +101675,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -101705,7 +101705,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -101781,7 +101781,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -101813,7 +101813,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -101942,7 +101942,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -102016,11 +102016,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -102030,7 +102030,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102042,11 +102042,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -102064,7 +102064,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -102163,11 +102163,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -102184,23 +102184,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -102225,7 +102225,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102233,7 +102233,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -102298,12 +102298,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -102393,7 +102393,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -102419,7 +102419,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102466,7 +102466,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -102581,7 +102581,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102589,7 +102589,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -102611,7 +102611,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102655,7 +102655,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102667,15 +102667,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -102693,7 +102693,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102729,7 +102729,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -102741,7 +102741,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -102773,7 +102773,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -102783,19 +102783,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -102809,7 +102809,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -102817,7 +102817,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -102826,17 +102826,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -102844,15 +102844,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -102968,19 +102968,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -102989,7 +102989,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -103009,11 +103009,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -103035,7 +103035,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -103054,14 +103054,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -103077,7 +103077,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -103101,7 +103101,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -103133,7 +103133,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -103147,12 +103147,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -103253,7 +103253,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -103274,7 +103274,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -103287,14 +103287,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -103441,11 +103441,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -103483,7 +103483,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -103531,7 +103531,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -103575,23 +103575,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -103625,11 +103625,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -103655,7 +103655,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -103667,15 +103667,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -103738,11 +103738,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -103780,7 +103780,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -103861,11 +103861,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -103903,7 +103903,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -103994,7 +103994,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -104421,7 +104421,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -104465,11 +104465,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -104526,7 +104526,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -104630,11 +104630,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -104672,7 +104672,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -104711,7 +104711,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -104803,7 +104803,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -104864,11 +104864,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -104925,7 +104925,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -104933,7 +104933,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -104980,7 +104980,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -105016,7 +105016,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -105056,7 +105056,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -105182,11 +105182,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -105239,7 +105239,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -105318,7 +105318,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -105356,7 +105356,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -105390,11 +105390,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -105404,7 +105404,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -105416,11 +105416,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -105438,7 +105438,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -105604,11 +105604,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -105625,23 +105625,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -105666,7 +105666,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -105674,7 +105674,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -105739,12 +105739,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -105834,7 +105834,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -105892,7 +105892,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -105939,7 +105939,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -106054,7 +106054,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106062,7 +106062,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -106076,7 +106076,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -106202,7 +106202,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106278,7 +106278,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106290,15 +106290,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -106316,7 +106316,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106352,7 +106352,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -106364,7 +106364,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -106396,7 +106396,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -106406,19 +106406,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -106432,7 +106432,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106488,11 +106488,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -106502,7 +106502,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106514,11 +106514,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -106536,7 +106536,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -106635,11 +106635,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -106656,23 +106656,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -106697,7 +106697,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106705,7 +106705,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -106770,12 +106770,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -106865,7 +106865,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -106891,7 +106891,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -106938,7 +106938,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -107053,7 +107053,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -107061,7 +107061,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -107083,7 +107083,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -107127,7 +107127,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -107139,15 +107139,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -107165,7 +107165,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -107201,7 +107201,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -107213,7 +107213,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -107245,7 +107245,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -107255,19 +107255,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -107281,7 +107281,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -107325,7 +107325,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -107355,7 +107355,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -107363,7 +107363,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -107442,15 +107442,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -107492,7 +107492,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -107516,11 +107516,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -107528,7 +107528,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -107570,7 +107570,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -107606,7 +107606,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -107670,7 +107670,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -107774,7 +107774,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -107824,7 +107824,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -107848,7 +107848,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -107878,16 +107878,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -107929,11 +107929,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -107943,7 +107943,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -107995,7 +107995,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -108025,7 +108025,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -108101,7 +108101,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -108318,7 +108318,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -108362,11 +108362,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -108423,7 +108423,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -108527,11 +108527,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -108569,7 +108569,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -108608,7 +108608,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -108700,7 +108700,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -108761,11 +108761,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -108822,7 +108822,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -108830,7 +108830,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -108877,7 +108877,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -108913,7 +108913,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -108953,7 +108953,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -109079,11 +109079,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -109136,7 +109136,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -109215,7 +109215,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -109253,7 +109253,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -109287,11 +109287,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -109301,7 +109301,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -109313,11 +109313,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -109335,7 +109335,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -109501,11 +109501,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -109522,23 +109522,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -109563,7 +109563,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -109571,7 +109571,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -109636,12 +109636,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -109731,7 +109731,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -109789,7 +109789,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -109836,7 +109836,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -109951,7 +109951,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -109959,7 +109959,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -109973,7 +109973,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -110099,7 +110099,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110175,7 +110175,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110187,15 +110187,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -110213,7 +110213,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110249,7 +110249,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -110261,7 +110261,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -110293,7 +110293,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -110303,19 +110303,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -110329,7 +110329,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110385,11 +110385,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -110399,7 +110399,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110411,11 +110411,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -110433,7 +110433,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -110532,11 +110532,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -110553,23 +110553,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -110594,7 +110594,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110602,7 +110602,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -110667,12 +110667,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -110762,7 +110762,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -110788,7 +110788,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110835,7 +110835,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -110950,7 +110950,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -110958,7 +110958,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -110980,7 +110980,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -111024,7 +111024,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -111036,15 +111036,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -111062,7 +111062,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -111098,7 +111098,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -111110,7 +111110,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -111142,7 +111142,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -111152,19 +111152,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -111178,7 +111178,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -111222,7 +111222,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -111252,7 +111252,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -111260,7 +111260,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -111339,15 +111339,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -111389,7 +111389,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -111413,11 +111413,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -111425,7 +111425,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -111467,7 +111467,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -111503,7 +111503,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -111567,7 +111567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -111671,7 +111671,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -111721,7 +111721,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -111745,7 +111745,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -111775,16 +111775,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -111826,11 +111826,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -111840,7 +111840,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -111892,7 +111892,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -111922,7 +111922,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -111998,7 +111998,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -112229,7 +112229,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -112273,11 +112273,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -112334,7 +112334,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -112438,11 +112438,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -112480,7 +112480,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -112519,7 +112519,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -112611,7 +112611,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -112672,11 +112672,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -112733,7 +112733,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -112741,7 +112741,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -112788,7 +112788,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -112824,7 +112824,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -112864,7 +112864,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -112990,11 +112990,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -113047,7 +113047,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -113126,7 +113126,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -113164,7 +113164,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -113198,11 +113198,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -113212,7 +113212,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -113224,11 +113224,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -113246,7 +113246,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -113412,11 +113412,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -113433,23 +113433,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -113474,7 +113474,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -113482,7 +113482,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -113547,12 +113547,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -113642,7 +113642,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -113700,7 +113700,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -113747,7 +113747,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -113862,7 +113862,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -113870,7 +113870,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -113884,7 +113884,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -114010,7 +114010,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114086,7 +114086,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114098,15 +114098,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -114124,7 +114124,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114160,7 +114160,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -114172,7 +114172,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -114204,7 +114204,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -114214,19 +114214,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -114240,7 +114240,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114296,11 +114296,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -114310,7 +114310,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114322,11 +114322,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -114344,7 +114344,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -114443,11 +114443,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -114464,23 +114464,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -114505,7 +114505,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114513,7 +114513,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -114578,12 +114578,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -114673,7 +114673,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -114699,7 +114699,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114746,7 +114746,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -114861,7 +114861,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114869,7 +114869,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -114891,7 +114891,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114935,7 +114935,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -114947,15 +114947,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -114973,7 +114973,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -115009,7 +115009,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -115021,7 +115021,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -115053,7 +115053,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -115063,19 +115063,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -115089,7 +115089,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -115133,7 +115133,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -115163,7 +115163,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -115171,7 +115171,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -115250,15 +115250,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -115300,7 +115300,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -115324,11 +115324,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -115336,7 +115336,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -115378,7 +115378,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -115414,7 +115414,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -115478,7 +115478,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -115582,7 +115582,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -115632,7 +115632,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -115656,7 +115656,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -115686,16 +115686,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -115737,11 +115737,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -115751,7 +115751,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -115803,7 +115803,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -115833,7 +115833,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -115909,7 +115909,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -116132,7 +116132,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -116337,7 +116337,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -116391,11 +116391,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -116433,7 +116433,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -116455,7 +116455,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"explicit"` - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. @@ -116477,7 +116477,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -116529,7 +116529,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -116541,7 +116541,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -116627,11 +116627,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -116669,7 +116669,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -116704,7 +116704,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -116937,7 +116937,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -116981,11 +116981,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -117036,11 +117036,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -117097,7 +117097,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -117105,7 +117105,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -117152,7 +117152,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -117188,7 +117188,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -117228,7 +117228,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -117330,11 +117330,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -117342,15 +117342,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -117416,7 +117416,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -117483,7 +117483,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -117552,7 +117552,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -117572,7 +117572,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -117584,15 +117584,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -117622,23 +117622,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -117654,11 +117654,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -117682,15 +117682,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -117712,11 +117712,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -117728,7 +117728,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -117752,11 +117752,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -117766,7 +117766,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -117778,11 +117778,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -117800,7 +117800,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -117966,11 +117966,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -117987,23 +117987,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -118028,7 +118028,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118036,7 +118036,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -118101,12 +118101,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -118196,7 +118196,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -118254,7 +118254,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118301,7 +118301,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -118416,7 +118416,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118424,7 +118424,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -118438,7 +118438,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -118564,7 +118564,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118640,7 +118640,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118652,15 +118652,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -118678,7 +118678,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118714,7 +118714,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -118726,7 +118726,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -118758,7 +118758,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -118768,19 +118768,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -118794,7 +118794,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118808,11 +118808,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -118824,7 +118824,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -118854,11 +118854,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -118868,7 +118868,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -118880,11 +118880,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -118902,7 +118902,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -119001,11 +119001,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -119022,23 +119022,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -119063,7 +119063,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -119071,7 +119071,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -119136,12 +119136,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -119231,7 +119231,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -119257,7 +119257,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -119304,7 +119304,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -119419,7 +119419,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -119427,7 +119427,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -119449,7 +119449,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -119493,7 +119493,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -119505,15 +119505,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -119531,7 +119531,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -119567,7 +119567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -119579,7 +119579,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -119611,7 +119611,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -119621,19 +119621,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -119647,7 +119647,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -119661,7 +119661,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -119710,7 +119710,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -119741,7 +119741,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -119753,7 +119753,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -119783,7 +119783,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -119791,7 +119791,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -119870,15 +119870,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -119920,7 +119920,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -119942,11 +119942,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -119960,11 +119960,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -119988,7 +119988,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -119996,7 +119996,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -120060,11 +120060,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -120088,11 +120088,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -120178,11 +120178,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -120228,11 +120228,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -120256,7 +120256,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -120284,11 +120284,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -120298,7 +120298,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -120346,11 +120346,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -120380,16 +120380,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -120450,7 +120450,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -120500,7 +120500,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -120542,7 +120542,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -120600,7 +120600,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -120867,11 +120867,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -120928,7 +120928,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -121015,7 +121015,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -121107,7 +121107,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -121168,11 +121168,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -121244,11 +121244,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -121299,7 +121299,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -121378,7 +121378,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -121416,7 +121416,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -121450,11 +121450,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -121464,7 +121464,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -121476,11 +121476,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -121498,7 +121498,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -121597,11 +121597,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -121618,23 +121618,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -121659,7 +121659,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -121667,7 +121667,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -121732,12 +121732,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -121827,7 +121827,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -121853,7 +121853,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -121900,7 +121900,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -122015,7 +122015,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122023,7 +122023,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -122045,7 +122045,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122089,7 +122089,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122101,15 +122101,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -122127,7 +122127,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122163,7 +122163,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -122175,7 +122175,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -122207,7 +122207,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -122217,19 +122217,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -122243,7 +122243,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122299,11 +122299,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -122313,7 +122313,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122325,11 +122325,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -122347,7 +122347,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -122446,11 +122446,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -122467,23 +122467,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -122508,7 +122508,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122516,7 +122516,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -122581,12 +122581,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -122676,7 +122676,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -122702,7 +122702,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122749,7 +122749,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -122864,7 +122864,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122872,7 +122872,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -122894,7 +122894,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122938,7 +122938,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -122950,15 +122950,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -122976,7 +122976,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -123012,7 +123012,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -123024,7 +123024,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -123056,7 +123056,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -123066,19 +123066,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -123092,7 +123092,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -123136,7 +123136,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -123166,7 +123166,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -123174,7 +123174,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -123253,15 +123253,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -123303,7 +123303,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -123327,11 +123327,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -123339,7 +123339,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -123381,7 +123381,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -123417,7 +123417,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -123481,7 +123481,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -123585,7 +123585,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -123635,7 +123635,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -123659,7 +123659,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -123689,16 +123689,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -123740,11 +123740,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -123754,7 +123754,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -123806,7 +123806,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -123836,7 +123836,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -123912,7 +123912,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -123944,7 +123944,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -124073,7 +124073,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -124147,11 +124147,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -124161,7 +124161,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124173,11 +124173,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -124195,7 +124195,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -124294,11 +124294,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -124315,23 +124315,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -124356,7 +124356,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124364,7 +124364,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -124429,12 +124429,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -124524,7 +124524,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -124550,7 +124550,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124597,7 +124597,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -124712,7 +124712,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124720,7 +124720,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -124742,7 +124742,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124786,7 +124786,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124798,15 +124798,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -124824,7 +124824,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124860,7 +124860,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -124872,7 +124872,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -124904,7 +124904,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -124914,19 +124914,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -124940,7 +124940,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -124948,7 +124948,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -124957,17 +124957,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -124975,15 +124975,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -125099,19 +125099,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -125120,7 +125120,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -125140,11 +125140,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -125166,7 +125166,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -125185,14 +125185,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -125208,7 +125208,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -125232,7 +125232,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -125264,7 +125264,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -125278,12 +125278,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -125384,7 +125384,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -125405,7 +125405,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -125418,14 +125418,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -125999,7 +125999,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -126051,7 +126051,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -126063,7 +126063,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -126149,11 +126149,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -126191,7 +126191,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -126226,7 +126226,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -126459,7 +126459,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -126503,11 +126503,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -126558,11 +126558,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -126619,7 +126619,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -126627,7 +126627,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -126674,7 +126674,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -126710,7 +126710,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -126750,7 +126750,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -126852,11 +126852,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -126864,15 +126864,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -126938,7 +126938,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -127005,7 +127005,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -127074,7 +127074,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -127094,7 +127094,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -127106,15 +127106,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -127144,23 +127144,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -127176,11 +127176,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -127204,15 +127204,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -127234,11 +127234,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -127250,7 +127250,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -127274,11 +127274,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -127288,7 +127288,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -127300,11 +127300,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -127322,7 +127322,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -127488,11 +127488,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -127509,23 +127509,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -127550,7 +127550,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -127558,7 +127558,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -127623,12 +127623,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -127718,7 +127718,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -127776,7 +127776,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -127823,7 +127823,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -127938,7 +127938,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -127946,7 +127946,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -127960,7 +127960,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -128086,7 +128086,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128162,7 +128162,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128174,15 +128174,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -128200,7 +128200,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128236,7 +128236,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -128248,7 +128248,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -128280,7 +128280,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -128290,19 +128290,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -128316,7 +128316,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128330,11 +128330,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -128346,7 +128346,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -128376,11 +128376,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -128390,7 +128390,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128402,11 +128402,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -128424,7 +128424,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -128523,11 +128523,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -128544,23 +128544,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -128585,7 +128585,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128593,7 +128593,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -128658,12 +128658,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -128753,7 +128753,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -128779,7 +128779,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128826,7 +128826,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -128941,7 +128941,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -128949,7 +128949,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -128971,7 +128971,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -129015,7 +129015,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -129027,15 +129027,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -129053,7 +129053,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -129089,7 +129089,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -129101,7 +129101,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -129133,7 +129133,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -129143,19 +129143,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -129169,7 +129169,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -129183,7 +129183,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -129232,7 +129232,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -129263,7 +129263,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -129275,7 +129275,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -129305,7 +129305,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -129313,7 +129313,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -129392,15 +129392,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -129442,7 +129442,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -129464,11 +129464,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -129482,11 +129482,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -129510,7 +129510,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -129518,7 +129518,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -129582,11 +129582,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -129610,11 +129610,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -129700,11 +129700,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -129750,11 +129750,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -129778,7 +129778,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -129806,11 +129806,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -129820,7 +129820,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -129868,11 +129868,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -129902,16 +129902,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -129972,7 +129972,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -130022,7 +130022,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -130064,7 +130064,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -130122,7 +130122,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -130389,11 +130389,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -130450,7 +130450,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -130537,7 +130537,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -130629,7 +130629,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -130690,11 +130690,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -130766,11 +130766,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -130821,7 +130821,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -130900,7 +130900,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -130938,7 +130938,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -130972,11 +130972,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -130986,7 +130986,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -130998,11 +130998,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -131020,7 +131020,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -131119,11 +131119,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -131140,23 +131140,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -131181,7 +131181,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131189,7 +131189,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -131254,12 +131254,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -131349,7 +131349,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -131375,7 +131375,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131422,7 +131422,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -131537,7 +131537,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131545,7 +131545,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -131567,7 +131567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131611,7 +131611,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131623,15 +131623,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -131649,7 +131649,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131685,7 +131685,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -131697,7 +131697,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -131729,7 +131729,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -131739,19 +131739,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -131765,7 +131765,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131821,11 +131821,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -131835,7 +131835,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -131847,11 +131847,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -131869,7 +131869,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -131968,11 +131968,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -131989,23 +131989,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -132030,7 +132030,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -132038,7 +132038,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -132103,12 +132103,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -132198,7 +132198,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -132224,7 +132224,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -132271,7 +132271,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -132386,7 +132386,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -132394,7 +132394,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -132416,7 +132416,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -132460,7 +132460,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -132472,15 +132472,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -132498,7 +132498,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -132534,7 +132534,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -132546,7 +132546,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -132578,7 +132578,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -132588,19 +132588,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -132614,7 +132614,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -132658,7 +132658,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -132688,7 +132688,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -132696,7 +132696,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -132775,15 +132775,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -132825,7 +132825,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -132849,11 +132849,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -132861,7 +132861,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -132903,7 +132903,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -132939,7 +132939,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -133003,7 +133003,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -133107,7 +133107,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -133157,7 +133157,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -133181,7 +133181,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -133211,16 +133211,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -133262,11 +133262,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -133276,7 +133276,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -133328,7 +133328,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -133358,7 +133358,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -133434,7 +133434,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -133466,7 +133466,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -133595,7 +133595,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -133669,11 +133669,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -133683,7 +133683,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -133695,11 +133695,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -133717,7 +133717,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -133816,11 +133816,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -133837,23 +133837,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -133878,7 +133878,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -133886,7 +133886,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -133951,12 +133951,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -134046,7 +134046,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -134072,7 +134072,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -134119,7 +134119,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -134234,7 +134234,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -134242,7 +134242,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -134264,7 +134264,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -134308,7 +134308,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -134320,15 +134320,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -134346,7 +134346,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -134382,7 +134382,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -134394,7 +134394,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -134426,7 +134426,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -134436,19 +134436,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -134462,7 +134462,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -134470,7 +134470,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -134479,17 +134479,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -134497,15 +134497,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -134621,19 +134621,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -134642,7 +134642,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -134662,11 +134662,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -134688,7 +134688,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -134707,14 +134707,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -134730,7 +134730,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -134754,7 +134754,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -134786,7 +134786,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -134800,12 +134800,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -134906,7 +134906,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -134927,7 +134927,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -134940,14 +134940,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -135141,7 +135141,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Emitted when an error occurs. - - `code: string` + - `code: string or null` The error code. @@ -135149,7 +135149,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The error message. - - `param: string` + - `param: string or null` The error parameter. @@ -136371,7 +136371,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -136392,7 +136392,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -136639,12 +136639,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"response.create"` - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `context_management: optional array of object { type, compact_threshold }` + - `context_management: optional array of object { type, compact_threshold } or null` Context management configuration for this request. @@ -136652,11 +136652,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The context management entry type. Currently only 'compaction' is supported. - - `compact_threshold: optional number` + - `compact_threshold: optional number or null` Token threshold at which compaction should be triggered for this entry. - - `conversation: optional string or ResponseConversationParam` + - `conversation: optional string or ResponseConversationParam or null` The conversation that this response belongs to. Items from this conversation are prepended to `input_items` for this response request. Input items and output items from this response are automatically added to this conversation after this response completes. @@ -136673,7 +136673,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation. - - `include: optional array of ResponseIncludable` + - `include: optional array of ResponseIncludable or null` Specify additional output data to include in the model response. Currently supported values are: @@ -136791,11 +136791,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -136833,7 +136833,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -136868,7 +136868,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -137101,7 +137101,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -137145,11 +137145,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -137200,11 +137200,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -137261,7 +137261,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -137269,7 +137269,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -137316,7 +137316,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -137352,7 +137352,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -137392,7 +137392,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -137494,11 +137494,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -137506,15 +137506,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -137580,7 +137580,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -137647,7 +137647,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -137716,7 +137716,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -137736,7 +137736,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -137748,15 +137748,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -137786,23 +137786,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -137818,11 +137818,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -137846,15 +137846,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -137876,11 +137876,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -137892,7 +137892,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -137916,11 +137916,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -137930,7 +137930,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -137942,11 +137942,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -137964,7 +137964,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -138130,11 +138130,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -138151,23 +138151,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -138192,7 +138192,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -138200,7 +138200,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -138265,12 +138265,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -138360,7 +138360,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -138418,7 +138418,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -138465,7 +138465,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -138580,7 +138580,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -138588,7 +138588,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -138602,7 +138602,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -138728,7 +138728,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -138804,7 +138804,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -138816,15 +138816,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -138842,7 +138842,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -138878,7 +138878,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -138890,7 +138890,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -138922,7 +138922,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -138932,19 +138932,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -138958,7 +138958,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -138972,11 +138972,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -138988,7 +138988,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -139018,11 +139018,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -139032,7 +139032,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139044,11 +139044,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -139066,7 +139066,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -139165,11 +139165,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -139186,23 +139186,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -139227,7 +139227,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139235,7 +139235,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -139300,12 +139300,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -139395,7 +139395,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -139421,7 +139421,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139468,7 +139468,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -139583,7 +139583,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139591,7 +139591,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -139613,7 +139613,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139657,7 +139657,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139669,15 +139669,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -139695,7 +139695,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139731,7 +139731,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -139743,7 +139743,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -139775,7 +139775,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -139785,19 +139785,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -139811,7 +139811,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -139825,7 +139825,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -139874,7 +139874,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -139905,7 +139905,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -139917,7 +139917,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -139947,7 +139947,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -139955,7 +139955,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -140034,15 +140034,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -140084,7 +140084,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -140106,11 +140106,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -140124,11 +140124,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -140152,7 +140152,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -140160,7 +140160,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -140224,11 +140224,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -140252,11 +140252,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -140342,11 +140342,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -140392,11 +140392,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -140420,7 +140420,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -140448,11 +140448,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -140462,7 +140462,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -140510,11 +140510,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -140544,16 +140544,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -140614,7 +140614,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -140664,7 +140664,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -140706,7 +140706,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -140764,7 +140764,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `instructions: optional string` + - `instructions: optional string or null` A system (or developer) message inserted into the model's context. @@ -140772,15 +140772,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `metadata: optional Metadata` + - `metadata: optional Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -140992,7 +140992,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"gpt-5.1-codex-max"` - - `moderation: optional object { model, policy }` + - `moderation: optional object { model, policy } or null` Configuration for running moderation on the input and output of this response. @@ -141000,11 +141000,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The moderation model to use for moderated completions, e.g. 'omni-moderation-latest'. - - `policy: optional object { input, output }` + - `policy: optional object { input, output } or null` The policy to apply to moderated response input and output. - - `input: optional object { mode }` + - `input: optional object { mode } or null` The moderation policy for the response input. @@ -141014,7 +141014,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"block"` - - `output: optional object { mode }` + - `output: optional object { mode } or null` The moderation policy for the response output. @@ -141024,17 +141024,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"block"` - - `parallel_tool_calls: optional boolean` + - `parallel_tool_calls: optional boolean or null` Whether to allow the model to run tool calls in parallel. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -141043,7 +141043,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -141063,11 +141063,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -141089,7 +141089,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -141108,14 +141108,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -141131,7 +141131,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -141155,7 +141155,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -141187,7 +141187,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -141201,12 +141201,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -141230,19 +141230,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"fast"` - - `store: optional boolean` + - `store: optional boolean or null` Whether to store the generated model response for later retrieval via API. - - `stream: optional boolean` + - `stream: optional boolean or null` If set to true, the model response data will be streamed to the client as it is generated using [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#Event_stream_format). See the [Streaming section below](/docs/api-reference/responses-streaming) for more information. - - `stream_options: optional object { include_obfuscation }` + - `stream_options: optional object { include_obfuscation } or null` Options for streaming responses. Only set this when you set `stream: true`. @@ -141256,7 +141256,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ false to optimize for bandwidth if you trust the network links between your application and the OpenAI API. - - `temperature: optional number` + - `temperature: optional number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -141321,7 +141321,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -141342,7 +141342,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -141479,7 +141479,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -141553,11 +141553,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -141567,7 +141567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -141579,11 +141579,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -141601,7 +141601,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -141700,11 +141700,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -141721,23 +141721,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -141762,7 +141762,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -141770,7 +141770,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -141835,12 +141835,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -141930,7 +141930,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -141956,7 +141956,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -142003,7 +142003,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -142118,7 +142118,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -142126,7 +142126,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -142148,7 +142148,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -142192,7 +142192,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -142204,15 +142204,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -142230,7 +142230,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -142266,7 +142266,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -142278,7 +142278,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -142310,7 +142310,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -142320,19 +142320,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -142346,7 +142346,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -142354,14 +142354,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `top_p: optional number` + - `top_p: optional number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -142370,7 +142370,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -142594,7 +142594,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Unix timestamp (in seconds) of when this Response was created. - - `error: ResponseError` + - `error: ResponseError or null` An error object returned when the model fails to generate a Response. @@ -142646,7 +142646,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A human-readable description of the error. - - `incomplete_details: object { reason }` + - `incomplete_details: object { reason } or null` Details about why the response is incomplete. @@ -142658,7 +142658,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"content_filter"` - - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` + - `instructions: string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` A system (or developer) message inserted into the model's context. @@ -142744,11 +142744,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -142786,7 +142786,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -142821,7 +142821,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -143054,7 +143054,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -143098,11 +143098,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -143153,11 +143153,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -143214,7 +143214,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -143222,7 +143222,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -143269,7 +143269,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -143305,7 +143305,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -143345,7 +143345,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -143447,11 +143447,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -143459,15 +143459,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -143533,7 +143533,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -143600,7 +143600,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -143669,7 +143669,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -143689,7 +143689,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -143701,15 +143701,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -143739,23 +143739,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -143771,11 +143771,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -143799,15 +143799,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -143829,11 +143829,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -143845,7 +143845,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -143869,11 +143869,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -143883,7 +143883,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -143895,11 +143895,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -143917,7 +143917,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -144083,11 +144083,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -144104,23 +144104,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -144145,7 +144145,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144153,7 +144153,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -144218,12 +144218,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -144313,7 +144313,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -144371,7 +144371,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144418,7 +144418,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -144533,7 +144533,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144541,7 +144541,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -144555,7 +144555,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -144681,7 +144681,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144757,7 +144757,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144769,15 +144769,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -144795,7 +144795,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144831,7 +144831,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -144843,7 +144843,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -144875,7 +144875,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -144885,19 +144885,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -144911,7 +144911,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144925,11 +144925,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -144941,7 +144941,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -144971,11 +144971,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -144985,7 +144985,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -144997,11 +144997,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -145019,7 +145019,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -145118,11 +145118,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -145139,23 +145139,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -145180,7 +145180,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -145188,7 +145188,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -145253,12 +145253,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -145348,7 +145348,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -145374,7 +145374,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -145421,7 +145421,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -145536,7 +145536,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -145544,7 +145544,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -145566,7 +145566,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -145610,7 +145610,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -145622,15 +145622,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -145648,7 +145648,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -145684,7 +145684,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -145696,7 +145696,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -145728,7 +145728,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -145738,19 +145738,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -145764,7 +145764,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -145778,7 +145778,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -145827,7 +145827,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -145858,7 +145858,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -145870,7 +145870,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -145900,7 +145900,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -145908,7 +145908,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -145987,15 +145987,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -146037,7 +146037,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -146059,11 +146059,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -146077,11 +146077,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -146105,7 +146105,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -146113,7 +146113,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -146177,11 +146177,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -146205,11 +146205,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -146295,11 +146295,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -146345,11 +146345,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -146373,7 +146373,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -146401,11 +146401,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -146415,7 +146415,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -146463,11 +146463,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -146497,16 +146497,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -146567,7 +146567,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -146617,7 +146617,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -146659,7 +146659,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -146717,7 +146717,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"program_output"` - - `metadata: Metadata` + - `metadata: Metadata or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -146984,11 +146984,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -147045,7 +147045,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -147132,7 +147132,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -147224,7 +147224,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -147285,11 +147285,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -147361,11 +147361,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -147416,7 +147416,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -147495,7 +147495,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -147533,7 +147533,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -147567,11 +147567,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -147581,7 +147581,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -147593,11 +147593,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -147615,7 +147615,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -147714,11 +147714,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -147735,23 +147735,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -147776,7 +147776,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -147784,7 +147784,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -147849,12 +147849,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -147944,7 +147944,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -147970,7 +147970,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148017,7 +148017,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -148132,7 +148132,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148140,7 +148140,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -148162,7 +148162,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148206,7 +148206,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148218,15 +148218,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -148244,7 +148244,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148280,7 +148280,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -148292,7 +148292,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -148324,7 +148324,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -148334,19 +148334,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -148360,7 +148360,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148416,11 +148416,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -148430,7 +148430,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148442,11 +148442,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -148464,7 +148464,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -148563,11 +148563,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -148584,23 +148584,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -148625,7 +148625,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148633,7 +148633,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -148698,12 +148698,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -148793,7 +148793,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -148819,7 +148819,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148866,7 +148866,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -148981,7 +148981,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -148989,7 +148989,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -149011,7 +149011,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -149055,7 +149055,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -149067,15 +149067,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -149093,7 +149093,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -149129,7 +149129,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -149141,7 +149141,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -149173,7 +149173,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -149183,19 +149183,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -149209,7 +149209,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -149253,7 +149253,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -149283,7 +149283,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -149291,7 +149291,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -149370,15 +149370,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -149420,7 +149420,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -149444,11 +149444,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -149456,7 +149456,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -149498,7 +149498,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -149534,7 +149534,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -149598,7 +149598,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -149702,7 +149702,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -149752,7 +149752,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -149776,7 +149776,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -149806,16 +149806,16 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -149857,11 +149857,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -149871,7 +149871,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -149923,7 +149923,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -149953,7 +149953,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -150029,7 +150029,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -150061,7 +150061,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether to allow the model to run tool calls in parallel. - - `temperature: number` + - `temperature: number or null` What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both. @@ -150190,7 +150190,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -150264,11 +150264,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -150278,7 +150278,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -150290,11 +150290,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -150312,7 +150312,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -150411,11 +150411,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -150432,23 +150432,23 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -150473,7 +150473,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -150481,7 +150481,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -150546,12 +150546,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -150641,7 +150641,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -150667,7 +150667,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -150714,7 +150714,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -150829,7 +150829,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -150837,7 +150837,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -150859,7 +150859,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -150903,7 +150903,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -150915,15 +150915,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -150941,7 +150941,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -150977,7 +150977,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -150989,7 +150989,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -151021,7 +151021,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -151031,19 +151031,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -151057,7 +151057,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -151065,7 +151065,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"programmatic"` - - `top_p: number` + - `top_p: number or null` An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability @@ -151074,17 +151074,17 @@ curl https://api.openai.com/v1/responses/resp_123 \ We generally recommend altering this or `temperature` but not both. - - `background: optional boolean` + - `background: optional boolean or null` Whether to run the model response in the background. [Learn more](/docs/guides/background). - - `completed_at: optional number` + - `completed_at: optional number or null` Unix timestamp (in seconds) of when this Response was completed. Only present when the status is `completed`. - - `conversation: optional object { id }` + - `conversation: optional object { id } or null` The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation. @@ -151092,15 +151092,15 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique ID of the conversation that this response was associated with. - - `max_output_tokens: optional number` + - `max_output_tokens: optional number or null` An upper bound for the number of tokens that can be generated for a response, including visible output tokens and [reasoning tokens](/docs/guides/reasoning). - - `max_tool_calls: optional number` + - `max_tool_calls: optional number or null` The maximum number of total calls to built-in tools that can be processed in a response. This maximum number applies across all built-in tool calls, not per individual tool. Any further attempts to call a tool by the model will be ignored. - - `moderation: optional object { input, output }` + - `moderation: optional object { input, output } or null` Moderation results for the response input and output, if moderated completions were requested. @@ -151216,19 +151216,19 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"error"` - - `output_text: optional string` + - `output_text: optional string or null` SDK-only convenience property that contains the aggregated text output from all `output_text` items in the `output` array, if any are present. Supported in the Python and JavaScript SDKs. - - `previous_response_id: optional string` + - `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. - - `prompt: optional ResponsePrompt` + - `prompt: optional ResponsePrompt or null` Reference to a prompt template and its variables. [Learn more](/docs/guides/text?api-mode=responses#reusable-prompts). @@ -151237,7 +151237,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The unique identifier of the prompt template to use. - - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile]` + - `variables: optional map[string or ResponseInputText or ResponseInputImage or ResponseInputFile] or null` Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other @@ -151257,11 +151257,11 @@ curl https://api.openai.com/v1/responses/resp_123 \ A file input to the model. - - `version: optional string` + - `version: optional string or null` Optional version of the prompt template. - - `prompt_cache_key: optional string` + - `prompt_cache_key: optional string or null` Used by OpenAI to cache responses for similar requests to optimize your cache hit rates. Replaces the `user` field. [Learn more](/docs/guides/prompt-caching). @@ -151283,7 +151283,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"30m"` - - `prompt_cache_retention: optional "in_memory" or "24h"` + - `prompt_cache_retention: optional "in_memory" or "24h" or null` Deprecated. Use `prompt_cache_options.ttl` instead. @@ -151302,14 +151302,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"24h"` - - `reasoning: optional Reasoning` + - `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -151325,7 +151325,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -151349,7 +151349,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -151381,7 +151381,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -151395,12 +151395,12 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"detailed"` - - `safety_identifier: optional string` + - `safety_identifier: optional string or null` A stable identifier used to help detect users of your application that may be violating OpenAI's usage policies. The IDs should be a string that uniquely identifies each user, with a maximum length of 64 characters. We recommend hashing their username or email address, in order to avoid sending us any identifying information. [Learn more](/docs/guides/safety-best-practices#safety-identifiers). - - `service_tier: optional "auto" or "default" or "flex" or 3 more` + - `service_tier: optional "auto" or "default" or "flex" or 3 more or null` Specifies the processing type used for serving the request. @@ -151501,7 +151501,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -151522,7 +151522,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -151535,14 +151535,14 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"high"` - - `top_logprobs: optional number` + - `top_logprobs: optional number or null` An integer between 0 and 20 specifying the maximum number of most likely tokens to return at each token position, each with an associated log probability. In some cases, the number of returned tokens may be fewer than requested. - - `truncation: optional "auto" or "disabled"` + - `truncation: optional "auto" or "disabled" or null` The truncation strategy to use for the model response. @@ -151736,7 +151736,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ Emitted when an error occurs. - - `code: string` + - `code: string or null` The error code. @@ -151744,7 +151744,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ The error message. - - `param: string` + - `param: string or null` The error parameter. @@ -153021,7 +153021,7 @@ curl https://api.openai.com/v1/responses/resp_123 \ - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -153216,11 +153216,11 @@ Returns a list of input items for a given response. - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -153258,7 +153258,7 @@ Returns a list of input items for a given response. The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -153486,7 +153486,7 @@ Returns a list of input items for a given response. - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -153530,11 +153530,11 @@ Returns a list of input items for a given response. - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -153585,11 +153585,11 @@ Returns a list of input items for a given response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -153646,7 +153646,7 @@ Returns a list of input items for a given response. The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -153654,7 +153654,7 @@ Returns a list of input items for a given response. A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -153701,7 +153701,7 @@ Returns a list of input items for a given response. - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -153737,7 +153737,7 @@ Returns a list of input items for a given response. The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -153777,7 +153777,7 @@ Returns a list of input items for a given response. The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -153903,11 +153903,11 @@ Returns a list of input items for a given response. The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -153971,7 +153971,7 @@ Returns a list of input items for a given response. - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -154046,7 +154046,7 @@ Returns a list of input items for a given response. - `"function_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -154126,7 +154126,7 @@ Returns a list of input items for a given response. - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -154172,7 +154172,7 @@ Returns a list of input items for a given response. Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -154210,7 +154210,7 @@ Returns a list of input items for a given response. The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -154244,11 +154244,11 @@ Returns a list of input items for a given response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -154258,7 +154258,7 @@ Returns a list of input items for a given response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -154270,11 +154270,11 @@ Returns a list of input items for a given response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -154292,7 +154292,7 @@ Returns a list of input items for a given response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -154458,11 +154458,11 @@ Returns a list of input items for a given response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -154479,23 +154479,23 @@ Returns a list of input items for a given response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -154520,7 +154520,7 @@ Returns a list of input items for a given response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -154528,7 +154528,7 @@ Returns a list of input items for a given response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -154593,12 +154593,12 @@ Returns a list of input items for a given response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -154688,7 +154688,7 @@ Returns a list of input items for a given response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -154746,7 +154746,7 @@ Returns a list of input items for a given response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -154793,7 +154793,7 @@ Returns a list of input items for a given response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -154908,7 +154908,7 @@ Returns a list of input items for a given response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -154916,7 +154916,7 @@ Returns a list of input items for a given response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -154930,7 +154930,7 @@ Returns a list of input items for a given response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -155056,7 +155056,7 @@ Returns a list of input items for a given response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155132,7 +155132,7 @@ Returns a list of input items for a given response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155144,15 +155144,15 @@ Returns a list of input items for a given response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -155170,7 +155170,7 @@ Returns a list of input items for a given response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155206,7 +155206,7 @@ Returns a list of input items for a given response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -155218,7 +155218,7 @@ Returns a list of input items for a given response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -155250,7 +155250,7 @@ Returns a list of input items for a given response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -155260,19 +155260,19 @@ Returns a list of input items for a given response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -155286,7 +155286,7 @@ Returns a list of input items for a given response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155342,11 +155342,11 @@ Returns a list of input items for a given response. The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -155356,7 +155356,7 @@ Returns a list of input items for a given response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155368,11 +155368,11 @@ Returns a list of input items for a given response. Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -155390,7 +155390,7 @@ Returns a list of input items for a given response. The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -155489,11 +155489,11 @@ Returns a list of input items for a given response. - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -155510,23 +155510,23 @@ Returns a list of input items for a given response. - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -155551,7 +155551,7 @@ Returns a list of input items for a given response. - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155559,7 +155559,7 @@ Returns a list of input items for a given response. - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -155624,12 +155624,12 @@ Returns a list of input items for a given response. Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -155719,7 +155719,7 @@ Returns a list of input items for a given response. An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -155745,7 +155745,7 @@ Returns a list of input items for a given response. - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155792,7 +155792,7 @@ Returns a list of input items for a given response. - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -155907,7 +155907,7 @@ Returns a list of input items for a given response. - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155915,7 +155915,7 @@ Returns a list of input items for a given response. - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -155937,7 +155937,7 @@ Returns a list of input items for a given response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155981,7 +155981,7 @@ Returns a list of input items for a given response. - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -155993,15 +155993,15 @@ Returns a list of input items for a given response. Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -156019,7 +156019,7 @@ Returns a list of input items for a given response. - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -156055,7 +156055,7 @@ Returns a list of input items for a given response. - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -156067,7 +156067,7 @@ Returns a list of input items for a given response. - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -156099,7 +156099,7 @@ Returns a list of input items for a given response. - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -156109,19 +156109,19 @@ Returns a list of input items for a given response. - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -156135,7 +156135,7 @@ Returns a list of input items for a given response. - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -156194,7 +156194,7 @@ Returns a list of input items for a given response. - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -156293,7 +156293,7 @@ Returns a list of input items for a given response. The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -156323,7 +156323,7 @@ Returns a list of input items for a given response. The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -156331,7 +156331,7 @@ Returns a list of input items for a given response. The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -156410,15 +156410,15 @@ Returns a list of input items for a given response. - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -156460,7 +156460,7 @@ Returns a list of input items for a given response. - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -156484,11 +156484,11 @@ Returns a list of input items for a given response. - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -156496,7 +156496,7 @@ Returns a list of input items for a given response. The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -156538,7 +156538,7 @@ Returns a list of input items for a given response. - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -156574,7 +156574,7 @@ Returns a list of input items for a given response. The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -156638,7 +156638,7 @@ Returns a list of input items for a given response. - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -156742,7 +156742,7 @@ Returns a list of input items for a given response. - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -156792,7 +156792,7 @@ Returns a list of input items for a given response. - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -156816,7 +156816,7 @@ Returns a list of input items for a given response. The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -156844,11 +156844,11 @@ Returns a list of input items for a given response. The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -156858,7 +156858,7 @@ Returns a list of input items for a given response. - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -156910,7 +156910,7 @@ Returns a list of input items for a given response. - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -156940,16 +156940,16 @@ Returns a list of input items for a given response. - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -157002,7 +157002,7 @@ Returns a list of input items for a given response. - `"custom_tool_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -157082,7 +157082,7 @@ Returns a list of input items for a given response. - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -157264,11 +157264,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -157306,7 +157306,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -157534,7 +157534,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -157578,11 +157578,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -157633,11 +157633,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -157694,7 +157694,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -157702,7 +157702,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -157749,7 +157749,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -157785,7 +157785,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -157825,7 +157825,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -157951,11 +157951,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -158019,7 +158019,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -158094,7 +158094,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"function_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -158174,7 +158174,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"function_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -158220,7 +158220,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ Arguments used for the tool search call. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -158258,7 +158258,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The unique ID of the tool search output item. - - `call_id: string` + - `call_id: string or null` The unique ID of the tool search call generated by the model. @@ -158292,11 +158292,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -158306,7 +158306,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -158318,11 +158318,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -158340,7 +158340,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -158506,11 +158506,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -158527,23 +158527,23 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -158568,7 +158568,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -158576,7 +158576,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -158641,12 +158641,12 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -158736,7 +158736,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -158794,7 +158794,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -158841,7 +158841,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -158956,7 +158956,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -158964,7 +158964,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -158978,7 +158978,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -159104,7 +159104,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159180,7 +159180,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159192,15 +159192,15 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -159218,7 +159218,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159254,7 +159254,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -159266,7 +159266,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -159298,7 +159298,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -159308,19 +159308,19 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -159334,7 +159334,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159390,11 +159390,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -159404,7 +159404,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159416,11 +159416,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -159438,7 +159438,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -159537,11 +159537,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -159558,23 +159558,23 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -159599,7 +159599,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159607,7 +159607,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -159672,12 +159672,12 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -159767,7 +159767,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -159793,7 +159793,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159840,7 +159840,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -159955,7 +159955,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -159963,7 +159963,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -159985,7 +159985,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -160029,7 +160029,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -160041,15 +160041,15 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -160067,7 +160067,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -160103,7 +160103,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -160115,7 +160115,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -160147,7 +160147,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -160157,19 +160157,19 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -160183,7 +160183,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -160242,7 +160242,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -160341,7 +160341,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -160371,7 +160371,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -160379,7 +160379,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -160458,15 +160458,15 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -160508,7 +160508,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -160532,11 +160532,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `commands: array of string` - - `max_output_length: number` + - `max_output_length: number or null` Optional maximum number of characters to return from each command. - - `timeout_ms: number` + - `timeout_ms: number or null` Optional timeout in milliseconds for the commands. @@ -160544,7 +160544,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The unique ID of the shell tool call generated by the model. - - `environment: ResponseLocalEnvironment or ResponseContainerReference` + - `environment: ResponseLocalEnvironment or ResponseContainerReference or null` Represents the use of a local environment to perform shell actions. @@ -160586,7 +160586,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"shell_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -160622,7 +160622,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The unique ID of the shell tool call generated by the model. - - `max_output_length: number` + - `max_output_length: number or null` The maximum length of the shell command output. This is generated by the model and should be passed back with the raw output. @@ -160686,7 +160686,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"shell_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -160790,7 +160790,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"apply_patch_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -160840,7 +160840,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"apply_patch_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -160864,7 +160864,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The ID of the entity that created this tool call output. - - `output: optional string` + - `output: optional string or null` Optional textual output returned by the apply patch tool. @@ -160892,11 +160892,11 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -160906,7 +160906,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -160958,7 +160958,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"mcp_approval_response"` - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -160988,16 +160988,16 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -161050,7 +161050,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"custom_tool_call"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -161130,7 +161130,7 @@ curl https://api.openai.com/v1/responses/resp_abc123/input_items \ - `"custom_tool_call_output"` - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -161188,7 +161188,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok ### Body Parameters -- `conversation: optional string or ResponseConversationParam` +- `conversation: optional string or ResponseConversationParam or null` The conversation that this response belongs to. Items from this conversation are prepended to `input_items` for this response request. Input items and output items from this response are automatically added to this conversation after this response completes. @@ -161205,7 +161205,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The unique ID of the conversation. -- `input: optional string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more` +- `input: optional string or array of EasyInputMessage or object { content, role, status, type } or ResponseOutputMessage or 29 more or null` Text, image, or file inputs to the model, used to generate a response @@ -161285,11 +161285,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"input_image"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. @@ -161327,7 +161327,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The content of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. @@ -161362,7 +161362,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"developer"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -161595,7 +161595,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"message"` - - `phase: optional "commentary" or "final_answer"` + - `phase: optional "commentary" or "final_answer" or null` Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`). For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend @@ -161639,11 +161639,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"file_search_call"` - - `results: optional array of object { attributes, file_id, filename, 2 more }` + - `results: optional array of object { attributes, file_id, filename, 2 more } or null` The results of the file search tool call. - - `attributes: optional map[string or number or boolean]` + - `attributes: optional map[string or number or boolean] or null` Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured @@ -161694,11 +161694,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. @@ -161755,7 +161755,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The y-coordinate where the click occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while clicking. @@ -161763,7 +161763,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok A double click action. - - `keys: array of string` + - `keys: array of string or null` The keys being held while double-clicking. @@ -161810,7 +161810,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"drag"` - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while dragging the mouse. @@ -161846,7 +161846,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The y-coordinate to move to. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while moving the mouse. @@ -161886,7 +161886,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The y-coordinate where the scroll occurred. - - `keys: optional array of string` + - `keys: optional array of string or null` The keys being held while scrolling. @@ -161988,11 +161988,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"computer_call_output"` - - `id: optional string` + - `id: optional string or null` The ID of the computer tool call output. - - `acknowledged_safety_checks: optional array of object { id, code, message }` + - `acknowledged_safety_checks: optional array of object { id, code, message } or null` The safety checks reported by the API that have been acknowledged by the developer. @@ -162000,15 +162000,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The ID of the pending safety check. - - `code: optional string` + - `code: optional string or null` The type of the pending safety check. - - `message: optional string` + - `message: optional string or null` Details about the pending safety check. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API. @@ -162074,7 +162074,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"open_page"` - - `url: optional string` + - `url: optional string or null` The URL opened by the model. @@ -162141,7 +162141,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The unique ID of the function tool call. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -162210,7 +162210,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"input_text"` - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -162230,7 +162230,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"input_image"` - - `detail: optional "low" or "high" or "auto" or "original"` + - `detail: optional "low" or "high" or "auto" or "original" or null` The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`. @@ -162242,15 +162242,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"original"` - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `image_url: optional string` + - `image_url: optional string or null` The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -162280,23 +162280,23 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` - - `file_data: optional string` + - `file_data: optional string or null` The base64-encoded data of the file to be sent to the model. - - `file_id: optional string` + - `file_id: optional string or null` The ID of the file to be sent to the model. - - `file_url: optional string` + - `file_url: optional string or null` The URL of the file to be sent to the model. - - `filename: optional string` + - `filename: optional string or null` The name of the file to be sent to the model. - - `prompt_cache_breakpoint: optional object { mode }` + - `prompt_cache_breakpoint: optional object { mode } or null` Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block. @@ -162312,11 +162312,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"function_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the function tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -162340,15 +162340,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"program"` - - `name: optional string` + - `name: optional string or null` The name of the tool that produced the output. - - `namespace: optional string` + - `namespace: optional string or null` The namespace of the tool that produced the output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API. @@ -162370,11 +162370,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"tool_search_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search call. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -162386,7 +162386,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search call. @@ -162410,11 +162410,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -162424,7 +162424,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -162436,11 +162436,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -162458,7 +162458,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -162624,11 +162624,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -162645,23 +162645,23 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -162686,7 +162686,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -162694,7 +162694,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -162759,12 +162759,12 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -162854,7 +162854,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -162912,7 +162912,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -162959,7 +162959,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -163074,7 +163074,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163082,7 +163082,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -163096,7 +163096,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the container. @@ -163222,7 +163222,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163298,7 +163298,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163310,15 +163310,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -163336,7 +163336,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163372,7 +163372,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -163384,7 +163384,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -163416,7 +163416,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -163426,19 +163426,19 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -163452,7 +163452,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163466,11 +163466,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"tool_search_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of this tool search output. - - `call_id: optional string` + - `call_id: optional string or null` The unique ID of the tool search call generated by the model. @@ -163482,7 +163482,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"client"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the tool search output. @@ -163512,11 +163512,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -163526,7 +163526,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163538,11 +163538,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -163560,7 +163560,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -163659,11 +163659,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -163680,23 +163680,23 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -163721,7 +163721,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163729,7 +163729,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -163794,12 +163794,12 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -163889,7 +163889,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -163915,7 +163915,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -163962,7 +163962,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -164077,7 +164077,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -164085,7 +164085,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -164107,7 +164107,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -164151,7 +164151,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -164163,15 +164163,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -164189,7 +164189,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -164225,7 +164225,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -164237,7 +164237,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -164269,7 +164269,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -164279,19 +164279,19 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -164305,7 +164305,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -164319,7 +164319,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"additional_tools"` - - `id: optional string` + - `id: optional string or null` The unique ID of this additional tools item. @@ -164368,7 +164368,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"reasoning_text"` - - `encrypted_content: optional string` + - `encrypted_content: optional string or null` The encrypted content of the reasoning item. This is populated by default for reasoning items returned by `POST /v1/responses` and WebSocket @@ -164399,7 +164399,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"compaction"` - - `id: optional string` + - `id: optional string or null` The ID of the compaction item. @@ -164411,7 +164411,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The unique ID of the image generation call. - - `result: string` + - `result: string or null` The generated image encoded in base64. @@ -164441,7 +164441,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The unique ID of the code interpreter tool call. - - `code: string` + - `code: string or null` The code to run, or null if not available. @@ -164449,7 +164449,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The ID of the container used to run the code. - - `outputs: array of object { logs, type } or object { type, url }` + - `outputs: array of object { logs, type } or object { type, url } or null` The outputs generated by the code interpreter, such as logs or images. Can be null if no outputs are available. @@ -164528,15 +164528,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"exec"` - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Optional timeout in milliseconds for the command. - - `user: optional string` + - `user: optional string or null` Optional user to run the command as. - - `working_directory: optional string` + - `working_directory: optional string or null` Optional working directory to run the command in. @@ -164578,7 +164578,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"local_shell_call_output"` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the item. One of `in_progress`, `completed`, or `incomplete`. @@ -164600,11 +164600,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Ordered shell commands for the execution environment to run. - - `max_output_length: optional number` + - `max_output_length: optional number or null` Maximum number of UTF-8 characters to capture from combined stdout and stderr output. - - `timeout_ms: optional number` + - `timeout_ms: optional number or null` Maximum wall-clock time in milliseconds to allow the shell commands to run. @@ -164618,11 +164618,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"shell_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -164646,7 +164646,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"program"` - - `environment: optional LocalEnvironment or ContainerReference` + - `environment: optional LocalEnvironment or ContainerReference or null` The environment to execute the shell commands in. @@ -164654,7 +164654,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `ContainerReference object { container_id, type }` - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call. One of `in_progress`, `completed`, or `incomplete`. @@ -164718,11 +164718,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"shell_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the shell tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -164746,11 +164746,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"program"` - - `max_output_length: optional number` + - `max_output_length: optional number or null` The maximum number of UTF-8 characters captured for this shell call's combined output. - - `status: optional "in_progress" or "completed" or "incomplete"` + - `status: optional "in_progress" or "completed" or "incomplete" or null` The status of the shell call output. @@ -164836,11 +164836,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"apply_patch_call"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -164886,11 +164886,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"apply_patch_call_output"` - - `id: optional string` + - `id: optional string or null` The unique ID of the apply patch tool call output. Populated when this item is returned via API. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -164914,7 +164914,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"program"` - - `output: optional string` + - `output: optional string or null` Optional human-readable log text from the apply patch tool (e.g., patch results or errors). @@ -164942,11 +164942,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The name of the tool. - - `annotations: optional unknown` + - `annotations: optional unknown or null` Additional annotations about the tool. - - `description: optional string` + - `description: optional string or null` The description of the tool. @@ -164956,7 +164956,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"mcp_list_tools"` - - `error: optional string` + - `error: optional string or null` Error message if the server could not list tools. @@ -165004,11 +165004,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"mcp_approval_response"` - - `id: optional string` + - `id: optional string or null` The unique ID of the approval response - - `reason: optional string` + - `reason: optional string or null` Optional reason for the decision. @@ -165038,16 +165038,16 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"mcp_call"` - - `approval_request_id: optional string` + - `approval_request_id: optional string or null` Unique identifier for the MCP tool call approval request. Include this value in a subsequent `mcp_approval_response` input to approve or reject the corresponding tool call. - - `error: optional string` + - `error: optional string or null` The error from the tool call, if any. - - `output: optional string` + - `output: optional string or null` The output from the tool call. @@ -165108,7 +165108,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The unique ID of the custom tool call output in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -165158,7 +165158,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The unique ID of the custom tool call in the OpenAI platform. - - `caller: optional object { type } or object { caller_id, type }` + - `caller: optional object { type } or object { caller_id, type } or null` The execution context that produced this tool call. @@ -165200,7 +165200,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The ID of the item to reference. - - `type: optional "item_reference"` + - `type: optional "item_reference" or null` The type of item to reference. Always `item_reference`. @@ -165258,16 +165258,16 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"program_output"` -- `instructions: optional string` +- `instructions: optional string or null` A system (or developer) message inserted into the model's context. When used along with `previous_response_id`, the instructions from a previous response will not be carried over to the next response. This makes it simple to swap out system (or developer) messages in new responses. -- `model: optional string` +- `model: optional string or null` Model ID used to generate the response, like `gpt-4o` or `o3`. OpenAI offers a wide range of models with different capabilities, performance characteristics, and price points. Refer to the [model guide](/docs/models) to browse and compare available models. -- `parallel_tool_calls: optional boolean` +- `parallel_tool_calls: optional boolean or null` Whether to allow the model to run tool calls in parallel. @@ -165285,15 +165285,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"pragmatic"` -- `previous_response_id: optional string` +- `previous_response_id: optional string or null` The unique ID of the previous response to the model. Use this to create multi-turn conversations. Learn more about [conversation state](/docs/guides/conversation-state). Cannot be used in conjunction with `conversation`. -- `reasoning: optional Reasoning` +- `reasoning: optional Reasoning or null` **gpt-5 and o-series models only** Configuration options for [reasoning models](https://platform.openai.com/docs/guides/reasoning). - - `context: optional "auto" or "current_turn" or "all_turns"` + - `context: optional "auto" or "current_turn" or "all_turns" or null` Controls which reasoning items are rendered back to the model on later turns. If omitted or set to `auto`, the model determines the context mode. The @@ -165309,7 +165309,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"all_turns"` - - `effort: optional ReasoningEffort` + - `effort: optional ReasoningEffort or null` Constrains effort on reasoning for reasoning models. Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`. @@ -165333,7 +165333,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"max"` - - `generate_summary: optional "auto" or "concise" or "detailed"` + - `generate_summary: optional "auto" or "concise" or "detailed" or null` **Deprecated:** use `summary` instead. @@ -165365,7 +165365,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"pro"` - - `summary: optional "auto" or "concise" or "detailed"` + - `summary: optional "auto" or "concise" or "detailed" or null` A summary of the reasoning performed by the model. This can be useful for debugging and understanding the model's reasoning process. @@ -165379,7 +165379,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"detailed"` -- `text: optional object { format, verbosity }` +- `text: optional object { format, verbosity } or null` Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: @@ -165439,7 +165439,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok A description of what the response format is for, used by the model to determine how to respond in the format. - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined @@ -165460,7 +165460,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"json_object"` - - `verbosity: optional "low" or "medium" or "high"` + - `verbosity: optional "low" or "medium" or "high" or null` Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. @@ -165473,7 +165473,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` -- `tool_choice: optional ToolChoiceOptions or ToolChoiceAllowed or ToolChoiceTypes or 6 more` +- `tool_choice: optional ToolChoiceOptions or ToolChoiceAllowed or ToolChoiceTypes or 6 more or null` Controls which tool the model should use, if any. @@ -165595,7 +165595,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"mcp"` - - `name: optional string` + - `name: optional string or null` The name of the tool to call on the server. @@ -165641,7 +165641,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"shell"` -- `tools: optional array of object { name, parameters, strict, 5 more } or object { type, vector_store_ids, filters, 2 more } or object { type } or 13 more` +- `tools: optional array of object { name, parameters, strict, 5 more } or object { type, vector_store_ids, filters, 2 more } or object { type } or 13 more or null` An array of tools the model may call while generating a response. You can specify which tool to use by setting the `tool_choice` parameter. @@ -165653,11 +165653,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The name of the function to call. - - `parameters: map[unknown]` + - `parameters: map[unknown] or null` A JSON schema object describing the parameters of the function. - - `strict: boolean` + - `strict: boolean or null` Whether strict parameter validation is enforced for this function tool. @@ -165667,7 +165667,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -165679,11 +165679,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this function is deferred and loaded via tool search. - - `description: optional string` + - `description: optional string or null` A description of the function. Used by the model to determine whether or not to call the function. - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON schema object describing the JSON value encoded in string outputs for this function. @@ -165701,7 +165701,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok The IDs of the vector stores to search. - - `filters: optional ComparisonFilter or CompoundFilter` + - `filters: optional ComparisonFilter or CompoundFilter or null` A filter to apply. @@ -165800,11 +165800,11 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"web_search_2025_08_26"` - - `filters: optional object { allowed_domains }` + - `filters: optional object { allowed_domains } or null` Filters for the search. - - `allowed_domains: optional array of string` + - `allowed_domains: optional array of string or null` Allowed domains for the search. If not provided, all domains are allowed. Subdomains of the provided domains are allowed as well. @@ -165821,23 +165821,23 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` - - `user_location: optional object { city, country, region, 2 more }` + - `user_location: optional object { city, country, region, 2 more } or null` The approximate location of the user. - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -165862,7 +165862,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"mcp"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -165870,7 +165870,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"programmatic"` - - `allowed_tools: optional array of string or object { read_only, tool_names }` + - `allowed_tools: optional array of string or object { read_only, tool_names } or null` List of allowed tool names or a filter object. @@ -165935,12 +165935,12 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this MCP tool is deferred and discovered via tool search. - - `headers: optional map[string]` + - `headers: optional map[string] or null` Optional HTTP headers to send to the MCP server. Use for authentication or other purposes. - - `require_approval: optional object { always, never } or "always" or "never"` + - `require_approval: optional object { always, never } or "always" or "never" or null` Specify which of the MCP server's tools require approval. @@ -166030,7 +166030,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok An optional list of uploaded files to make available to your code. - - `memory_limit: optional "1g" or "4g" or "16g" or "64g"` + - `memory_limit: optional "1g" or "4g" or "16g" or "64g" or null` The memory limit for the code interpreter container. @@ -166056,7 +166056,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"code_interpreter"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -166103,7 +166103,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"auto"` - - `input_fidelity: optional "high" or "low"` + - `input_fidelity: optional "high" or "low" or null` Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`. @@ -166218,7 +166218,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"shell"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -166226,7 +166226,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"programmatic"` - - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference` + - `environment: optional ContainerAuto or LocalEnvironment or ContainerReference or null` - `ContainerAuto object { type, file_ids, memory_limit, 2 more }` @@ -166248,7 +166248,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -166292,7 +166292,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"function"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -166304,15 +166304,15 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok Whether this function should be deferred and discovered via tool search. - - `description: optional string` + - `description: optional string or null` - - `output_schema: optional map[unknown]` + - `output_schema: optional map[unknown] or null` A JSON Schema describing the JSON value encoded in string outputs for this function tool. This does not describe content-array outputs. - - `parameters: optional unknown` + - `parameters: optional unknown or null` - - `strict: optional boolean` + - `strict: optional boolean or null` Whether to enforce strict parameter validation. If omitted, Responses attempts to use strict validation when the schema is compatible, and falls back to non-strict validation otherwise. @@ -166330,7 +166330,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"custom"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s). @@ -166366,7 +166366,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"tool_search"` - - `description: optional string` + - `description: optional string or null` Description shown to the model for a client-executed tool search tool. @@ -166378,7 +166378,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"client"` - - `parameters: optional unknown` + - `parameters: optional unknown or null` Parameter schema for a client-executed tool search tool. @@ -166410,7 +166410,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"high"` - - `user_location: optional object { type, city, country, 2 more }` + - `user_location: optional object { type, city, country, 2 more } or null` The user's location. @@ -166420,19 +166420,19 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"approximate"` - - `city: optional string` + - `city: optional string or null` Free text input for the city of the user, e.g. `San Francisco`. - - `country: optional string` + - `country: optional string or null` The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`. - - `region: optional string` + - `region: optional string or null` Free text input for the region of the user, e.g. `California`. - - `timezone: optional string` + - `timezone: optional string or null` The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`. @@ -166446,7 +166446,7 @@ Returns an object with `object` set to `response.input_tokens` and an `input_tok - `"apply_patch"` - - `allowed_callers: optional array of "direct" or "programmatic"` + - `allowed_callers: optional array of "direct" or "programmatic" or null` The tool invocation context(s).