Assistants

Create assistant

post /assistants

Create an assistant with a model and instructions.

Body Parameters

• model: string or "gpt-5" or "gpt-5-mini" or "gpt-5-nano" or 39 more

  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

  • string

  • AssistantSupportedModels = "gpt-5" or "gpt-5-mini" or "gpt-5-nano" or 39 more

    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

    • "gpt-5"

    • "gpt-5-mini"

    • "gpt-5-nano"

    • "gpt-5-2025-08-07"

    • "gpt-5-mini-2025-08-07"

    • "gpt-5-nano-2025-08-07"

    • "gpt-4.1"

    • "gpt-4.1-mini"

    • "gpt-4.1-nano"

    • "gpt-4.1-2025-04-14"

    • "gpt-4.1-mini-2025-04-14"

    • "gpt-4.1-nano-2025-04-14"

    • "o3-mini"

    • "o3-mini-2025-01-31"

    • "o1"

    • "o1-2024-12-17"

    • "gpt-4o"

    • "gpt-4o-2024-11-20"

    • "gpt-4o-2024-08-06"

    • "gpt-4o-2024-05-13"

    • "gpt-4o-mini"

    • "gpt-4o-mini-2024-07-18"

    • "gpt-4.5-preview"

    • "gpt-4.5-preview-2025-02-27"

    • "gpt-4-turbo"

    • "gpt-4-turbo-2024-04-09"

    • "gpt-4-0125-preview"

    • "gpt-4-turbo-preview"

    • "gpt-4-1106-preview"

    • "gpt-4-vision-preview"

    • "gpt-4"

    • "gpt-4-0314"

    • "gpt-4-0613"

    • "gpt-4-32k"

    • "gpt-4-32k-0314"

    • "gpt-4-32k-0613"

    • "gpt-3.5-turbo"

    • "gpt-3.5-turbo-16k"

    • "gpt-3.5-turbo-0613"

    • "gpt-3.5-turbo-1106"

    • "gpt-3.5-turbo-0125"

    • "gpt-3.5-turbo-16k-0613"

• description: optional string or null

  The description of the assistant. The maximum length is 512 characters.

• instructions: optional string or null

  The system instructions that the assistant uses. The maximum length is 256,000 characters.

• 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 format, and querying for objects via API or the dashboard.

  Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

• name: optional string or null

  The name of the assistant. The maximum length is 256 characters.

• reasoning_effort: optional ReasoningEffort or null

  Constrains effort on reasoning for reasoning models. Currently supported values are none, minimal, low, medium, high, xhigh, and max. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response. Not all reasoning models support every value. See the reasoning guide for model-specific support.

  • "none"

  • "minimal"

  • "low"

  • "medium"

  • "high"

  • "xhigh"

  • "max"

• response_format: optional AssistantResponseFormatOption or null

  Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

  Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

  Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

  Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

  • "auto"

    auto is the default value

    • "auto"

  • ResponseFormatText object { type }

    Default response format. Used to generate text responses.

    • type: "text"

      The type of response format being defined. Always text.

      • "text"

  • ResponseFormatJSONObject object { type }

    JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

    • type: "json_object"

      The type of response format being defined. Always json_object.

      • "json_object"

  • ResponseFormatJSONSchema object { json_schema, type }

    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

    • json_schema: object { name, description, schema, strict }

      Structured Outputs configuration options, including a JSON Schema.

      • name: string

        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

      • description: optional string

        A description of what the response format is for, used by the model to determine how to respond in the format.

      • schema: optional map[unknown]

        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

      • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

    • type: "json_schema"

      The type of response format being defined. Always json_schema.

      • "json_schema"

• 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.

• tool_resources: optional object { code_interpreter, file_search } or null

  A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

  • code_interpreter: optional object { file_ids }

    • file_ids: optional array of string

      A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.

  • file_search: optional object { vector_store_ids, vector_stores }

    • vector_store_ids: optional array of string

      The vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

    • vector_stores: optional array of object { chunking_strategy, file_ids, metadata }

      A helper to create a vector store with file_ids and attach it to this assistant. There can be a maximum of 1 vector store attached to the assistant.

      • chunking_strategy: optional object { type } or object { static, type }

        The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.

        • Auto object { type }

          The default strategy. This strategy currently uses a max_chunk_size_tokens of 800 and chunk_overlap_tokens of 400.

          • type: "auto"

            Always auto.

            • "auto"

        • Static object { static, type }

          • static: object { chunk_overlap_tokens, max_chunk_size_tokens }

            • chunk_overlap_tokens: number

              The number of tokens that overlap between chunks. The default value is 400.

              Note that the overlap must not exceed half of max_chunk_size_tokens.

            • max_chunk_size_tokens: number

              The maximum number of tokens in each chunk. The default value is 800. The minimum value is 100 and the maximum value is 4096.

          • type: "static"

            Always static.

            • "static"

      • file_ids: optional array of string

        A list of file IDs to add to the vector store. For vector stores created before Nov 2025, there can be a maximum of 10,000 files in a vector store. For vector stores created starting in Nov 2025, the limit is 100,000,000 files.

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

• tools: optional array of CodeInterpreterTool or FileSearchTool or FunctionTool

  A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter, file_search, or function.

  • CodeInterpreterTool object { type }

    • type: "code_interpreter"

      The type of tool being defined: code_interpreter

      • "code_interpreter"

  • FileSearchTool object { type, file_search }

    • type: "file_search"

      The type of tool being defined: file_search

      • "file_search"

    • file_search: optional object { max_num_results, ranking_options }

      Overrides for the file search tool.

      • max_num_results: optional number

        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

      • ranking_options: optional object { score_threshold, ranker }

        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

        See the file search tool documentation for more information.

        • score_threshold: number

          The score threshold for the file search. All values must be a floating point number between 0 and 1.

        • ranker: optional "auto" or "default_2024_08_21"

          The ranker to use for the file search. If not specified will use the auto ranker.

          • "auto"

          • "default_2024_08_21"

  • FunctionTool object { function, type }

    • function: FunctionDefinition

      • name: string

        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

      • description: optional string

        A description of what the function does, used by the model to choose when and how to call the function.

      • parameters: optional FunctionParameters

        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

        Omitting parameters defines a function with an empty parameter list.

      • strict: optional boolean or null

        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

    • type: "function"

      The type of tool being defined: function

      • "function"

• 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 mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

  We generally recommend altering this or temperature but not both.

Returns

• Assistant object { id, created_at, description, 10 more }

  Represents an assistant that can call the model and use tools.

  • id: string

    The identifier, which can be referenced in API endpoints.

  • created_at: number

    The Unix timestamp (in seconds) for when the assistant was created.

  • description: string or null

    The description of the assistant. The maximum length is 512 characters.

  • instructions: string or null

    The system instructions that the assistant uses. The maximum length is 256,000 characters.

  • 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 format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • model: string

    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

  • name: string or null

    The name of the assistant. The maximum length is 256 characters.

  • object: "assistant"

    The object type, which is always assistant.

    • "assistant"

  • tools: array of CodeInterpreterTool or FileSearchTool or FunctionTool

    A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter, file_search, or function.

    • CodeInterpreterTool object { type }

      • type: "code_interpreter"

        The type of tool being defined: code_interpreter

        • "code_interpreter"

    • FileSearchTool object { type, file_search }

      • type: "file_search"

        The type of tool being defined: file_search

        • "file_search"

      • file_search: optional object { max_num_results, ranking_options }

        Overrides for the file search tool.

        • max_num_results: optional number

          The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

          Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

        • ranking_options: optional object { score_threshold, ranker }

          The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

          See the file search tool documentation for more information.

          • score_threshold: number

            The score threshold for the file search. All values must be a floating point number between 0 and 1.

          • ranker: optional "auto" or "default_2024_08_21"

            The ranker to use for the file search. If not specified will use the auto ranker.

            • "auto"

            • "default_2024_08_21"

    • FunctionTool object { function, type }

      • function: FunctionDefinition

        • name: string

          The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the function does, used by the model to choose when and how to call the function.

        • parameters: optional FunctionParameters

          The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

          Omitting parameters defines a function with an empty parameter list.

        • strict: optional boolean or null

          Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

      • type: "function"

        The type of tool being defined: function

        • "function"

  • response_format: optional AssistantResponseFormatOption or null

    Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

    Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

    Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

    Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

    • "auto"

      auto is the default value

      • "auto"

    • ResponseFormatText object { type }

      Default response format. Used to generate text responses.

      • type: "text"

        The type of response format being defined. Always text.

        • "text"

    • ResponseFormatJSONObject object { type }

      JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

      • type: "json_object"

        The type of response format being defined. Always json_object.

        • "json_object"

    • ResponseFormatJSONSchema object { json_schema, type }

      JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

      • json_schema: object { name, description, schema, strict }

        Structured Outputs configuration options, including a JSON Schema.

        • name: string

          The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the response format is for, used by the model to determine how to respond in the format.

        • schema: optional map[unknown]

          The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

        • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

      • type: "json_schema"

        The type of response format being defined. Always json_schema.

        • "json_schema"

  • 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.

  • tool_resources: optional object { code_interpreter, file_search } or null

    A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

    • code_interpreter: optional object { file_ids }

      • file_ids: optional array of string

        A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

    • file_search: optional object { vector_store_ids }

      • vector_store_ids: optional array of string

        The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

  • 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 mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    We generally recommend altering this or temperature but not both.

Example

curl https://api.openai.com/v1/assistants \
    -H 'Content-Type: application/json' \
    -H 'OpenAI-Beta: assistants=v2' \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d '{
          "model": "gpt-4o",
          "temperature": 1,
          "top_p": 1
        }'

Response

{
  "id": "id",
  "created_at": 0,
  "description": "description",
  "instructions": "instructions",
  "metadata": {
    "foo": "string"
  },
  "model": "model",
  "name": "name",
  "object": "assistant",
  "tools": [
    {
      "type": "code_interpreter"
    }
  ],
  "response_format": "auto",
  "temperature": 1,
  "tool_resources": {
    "code_interpreter": {
      "file_ids": [
        "string"
      ]
    },
    "file_search": {
      "vector_store_ids": [
        "string"
      ]
    }
  },
  "top_p": 1
}

Code Interpreter

curl "https://api.openai.com/v1/assistants" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "OpenAI-Beta: assistants=v2" \
  -d '{
    "instructions": "You are a personal math tutor. When asked a question, write and run Python code to answer the question.",
    "name": "Math Tutor",
    "tools": [{"type": "code_interpreter"}],
    "model": "gpt-4o"
  }'

Response

{
  "id": "asst_abc123",
  "object": "assistant",
  "created_at": 1698984975,
  "name": "Math Tutor",
  "description": null,
  "model": "gpt-4o",
  "instructions": "You are a personal math tutor. When asked a question, write and run Python code to answer the question.",
  "tools": [
    {
      "type": "code_interpreter"
    }
  ],
  "metadata": {},
  "top_p": 1.0,
  "temperature": 1.0,
  "response_format": "auto"
}

Files

curl https://api.openai.com/v1/assistants \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "OpenAI-Beta: assistants=v2" \
  -d '{
    "instructions": "You are an HR bot, and you have access to files to answer employee questions about company policies.",
    "tools": [{"type": "file_search"}],
    "tool_resources": {"file_search": {"vector_store_ids": ["vs_123"]}},
    "model": "gpt-4o"
  }'

Response

{
  "id": "asst_abc123",
  "object": "assistant",
  "created_at": 1699009403,
  "name": "HR Helper",
  "description": null,
  "model": "gpt-4o",
  "instructions": "You are an HR bot, and you have access to files to answer employee questions about company policies.",
  "tools": [
    {
      "type": "file_search"
    }
  ],
  "tool_resources": {
    "file_search": {
      "vector_store_ids": ["vs_123"]
    }
  },
  "metadata": {},
  "top_p": 1.0,
  "temperature": 1.0,
  "response_format": "auto"
}

Delete assistant

delete /assistants/{assistant_id}

Delete an assistant.

Path Parameters

• assistant_id: string

Returns

• AssistantDeleted object { id, deleted, object }

  • id: string

  • deleted: boolean

  • object: "assistant.deleted"

    • "assistant.deleted"

Example

curl https://api.openai.com/v1/assistants/$ASSISTANT_ID \
    -X DELETE \
    -H 'OpenAI-Beta: assistants=v2' \
    -H "Authorization: Bearer $OPENAI_API_KEY"

Response

{
  "id": "id",
  "deleted": true,
  "object": "assistant.deleted"
}

Example

curl https://api.openai.com/v1/assistants/asst_abc123 \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "OpenAI-Beta: assistants=v2" \
  -X DELETE

Response

{
  "id": "asst_abc123",
  "object": "assistant.deleted",
  "deleted": true
}

List assistants

get /assistants

Returns a list of assistants.

Query Parameters

• after: optional string

  A cursor for use in pagination. after is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include after=obj_foo in order to fetch the next page of the list.

• before: optional string

  A cursor for use in pagination. before is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, starting with obj_foo, your subsequent call can include before=obj_foo in order to fetch the previous page of the list.

• limit: optional number

  A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.

• order: optional "asc" or "desc"

  Sort order by the created_at timestamp of the objects. asc for ascending order and desc for descending order.

  • "asc"

  • "desc"

Returns

• data: array of Assistant

  • id: string

    The identifier, which can be referenced in API endpoints.

  • created_at: number

    The Unix timestamp (in seconds) for when the assistant was created.

  • description: string or null

    The description of the assistant. The maximum length is 512 characters.

  • instructions: string or null

    The system instructions that the assistant uses. The maximum length is 256,000 characters.

  • 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 format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • model: string

    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

  • name: string or null

    The name of the assistant. The maximum length is 256 characters.

  • object: "assistant"

    The object type, which is always assistant.

    • "assistant"

  • tools: array of CodeInterpreterTool or FileSearchTool or FunctionTool

    A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter, file_search, or function.

    • CodeInterpreterTool object { type }

      • type: "code_interpreter"

        The type of tool being defined: code_interpreter

        • "code_interpreter"

    • FileSearchTool object { type, file_search }

      • type: "file_search"

        The type of tool being defined: file_search

        • "file_search"

      • file_search: optional object { max_num_results, ranking_options }

        Overrides for the file search tool.

        • max_num_results: optional number

          The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

          Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

        • ranking_options: optional object { score_threshold, ranker }

          The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

          See the file search tool documentation for more information.

          • score_threshold: number

            The score threshold for the file search. All values must be a floating point number between 0 and 1.

          • ranker: optional "auto" or "default_2024_08_21"

            The ranker to use for the file search. If not specified will use the auto ranker.

            • "auto"

            • "default_2024_08_21"

    • FunctionTool object { function, type }

      • function: FunctionDefinition

        • name: string

          The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the function does, used by the model to choose when and how to call the function.

        • parameters: optional FunctionParameters

          The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

          Omitting parameters defines a function with an empty parameter list.

        • strict: optional boolean or null

          Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

      • type: "function"

        The type of tool being defined: function

        • "function"

  • response_format: optional AssistantResponseFormatOption or null

    Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

    Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

    Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

    Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

    • "auto"

      auto is the default value

      • "auto"

    • ResponseFormatText object { type }

      Default response format. Used to generate text responses.

      • type: "text"

        The type of response format being defined. Always text.

        • "text"

    • ResponseFormatJSONObject object { type }

      JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

      • type: "json_object"

        The type of response format being defined. Always json_object.

        • "json_object"

    • ResponseFormatJSONSchema object { json_schema, type }

      JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

      • json_schema: object { name, description, schema, strict }

        Structured Outputs configuration options, including a JSON Schema.

        • name: string

          The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the response format is for, used by the model to determine how to respond in the format.

        • schema: optional map[unknown]

          The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

        • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

      • type: "json_schema"

        The type of response format being defined. Always json_schema.

        • "json_schema"

  • 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.

  • tool_resources: optional object { code_interpreter, file_search } or null

    A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

    • code_interpreter: optional object { file_ids }

      • file_ids: optional array of string

        A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

    • file_search: optional object { vector_store_ids }

      • vector_store_ids: optional array of string

        The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

  • 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 mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    We generally recommend altering this or temperature but not both.

• first_id: string

• has_more: boolean

• last_id: string

• object: string

Example

curl https://api.openai.com/v1/assistants \
    -H 'OpenAI-Beta: assistants=v2' \
    -H "Authorization: Bearer $OPENAI_API_KEY"

Response

{
  "data": [
    {
      "id": "id",
      "created_at": 0,
      "description": "description",
      "instructions": "instructions",
      "metadata": {
        "foo": "string"
      },
      "model": "model",
      "name": "name",
      "object": "assistant",
      "tools": [
        {
          "type": "code_interpreter"
        }
      ],
      "response_format": "auto",
      "temperature": 1,
      "tool_resources": {
        "code_interpreter": {
          "file_ids": [
            "string"
          ]
        },
        "file_search": {
          "vector_store_ids": [
            "string"
          ]
        }
      },
      "top_p": 1
    }
  ],
  "first_id": "asst_abc123",
  "has_more": false,
  "last_id": "asst_abc456",
  "object": "list"
}

Example

curl "https://api.openai.com/v1/assistants?order=desc&limit=20" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "OpenAI-Beta: assistants=v2"

Response

{
  "object": "list",
  "data": [
    {
      "id": "asst_abc123",
      "object": "assistant",
      "created_at": 1698982736,
      "name": "Coding Tutor",
      "description": null,
      "model": "gpt-4o",
      "instructions": "You are a helpful assistant designed to make me better at coding!",
      "tools": [],
      "tool_resources": {},
      "metadata": {},
      "top_p": 1.0,
      "temperature": 1.0,
      "response_format": "auto"
    },
    {
      "id": "asst_abc456",
      "object": "assistant",
      "created_at": 1698982718,
      "name": "My Assistant",
      "description": null,
      "model": "gpt-4o",
      "instructions": "You are a helpful assistant designed to make me better at coding!",
      "tools": [],
      "tool_resources": {},
      "metadata": {},
      "top_p": 1.0,
      "temperature": 1.0,
      "response_format": "auto"
    },
    {
      "id": "asst_abc789",
      "object": "assistant",
      "created_at": 1698982643,
      "name": null,
      "description": null,
      "model": "gpt-4o",
      "instructions": null,
      "tools": [],
      "tool_resources": {},
      "metadata": {},
      "top_p": 1.0,
      "temperature": 1.0,
      "response_format": "auto"
    }
  ],
  "first_id": "asst_abc123",
  "last_id": "asst_abc789",
  "has_more": false
}

Retrieve assistant

get /assistants/{assistant_id}

Retrieves an assistant.

Path Parameters

• assistant_id: string

Returns

• Assistant object { id, created_at, description, 10 more }

  Represents an assistant that can call the model and use tools.

  • id: string

    The identifier, which can be referenced in API endpoints.

  • created_at: number

    The Unix timestamp (in seconds) for when the assistant was created.

  • description: string or null

    The description of the assistant. The maximum length is 512 characters.

  • instructions: string or null

    The system instructions that the assistant uses. The maximum length is 256,000 characters.

  • 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 format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • model: string

    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

  • name: string or null

    The name of the assistant. The maximum length is 256 characters.

  • object: "assistant"

    The object type, which is always assistant.

    • "assistant"

  • tools: array of CodeInterpreterTool or FileSearchTool or FunctionTool

    A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter, file_search, or function.

    • CodeInterpreterTool object { type }

      • type: "code_interpreter"

        The type of tool being defined: code_interpreter

        • "code_interpreter"

    • FileSearchTool object { type, file_search }

      • type: "file_search"

        The type of tool being defined: file_search

        • "file_search"

      • file_search: optional object { max_num_results, ranking_options }

        Overrides for the file search tool.

        • max_num_results: optional number

          The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

          Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

        • ranking_options: optional object { score_threshold, ranker }

          The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

          See the file search tool documentation for more information.

          • score_threshold: number

            The score threshold for the file search. All values must be a floating point number between 0 and 1.

          • ranker: optional "auto" or "default_2024_08_21"

            The ranker to use for the file search. If not specified will use the auto ranker.

            • "auto"

            • "default_2024_08_21"

    • FunctionTool object { function, type }

      • function: FunctionDefinition

        • name: string

          The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the function does, used by the model to choose when and how to call the function.

        • parameters: optional FunctionParameters

          The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

          Omitting parameters defines a function with an empty parameter list.

        • strict: optional boolean or null

          Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

      • type: "function"

        The type of tool being defined: function

        • "function"

  • response_format: optional AssistantResponseFormatOption or null

    Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

    Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

    Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

    Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

    • "auto"

      auto is the default value

      • "auto"

    • ResponseFormatText object { type }

      Default response format. Used to generate text responses.

      • type: "text"

        The type of response format being defined. Always text.

        • "text"

    • ResponseFormatJSONObject object { type }

      JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

      • type: "json_object"

        The type of response format being defined. Always json_object.

        • "json_object"

    • ResponseFormatJSONSchema object { json_schema, type }

      JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

      • json_schema: object { name, description, schema, strict }

        Structured Outputs configuration options, including a JSON Schema.

        • name: string

          The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the response format is for, used by the model to determine how to respond in the format.

        • schema: optional map[unknown]

          The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

        • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

      • type: "json_schema"

        The type of response format being defined. Always json_schema.

        • "json_schema"

  • 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.

  • tool_resources: optional object { code_interpreter, file_search } or null

    A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

    • code_interpreter: optional object { file_ids }

      • file_ids: optional array of string

        A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

    • file_search: optional object { vector_store_ids }

      • vector_store_ids: optional array of string

        The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

  • 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 mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    We generally recommend altering this or temperature but not both.

Example

curl https://api.openai.com/v1/assistants/$ASSISTANT_ID \
    -H 'OpenAI-Beta: assistants=v2' \
    -H "Authorization: Bearer $OPENAI_API_KEY"

Response

{
  "id": "id",
  "created_at": 0,
  "description": "description",
  "instructions": "instructions",
  "metadata": {
    "foo": "string"
  },
  "model": "model",
  "name": "name",
  "object": "assistant",
  "tools": [
    {
      "type": "code_interpreter"
    }
  ],
  "response_format": "auto",
  "temperature": 1,
  "tool_resources": {
    "code_interpreter": {
      "file_ids": [
        "string"
      ]
    },
    "file_search": {
      "vector_store_ids": [
        "string"
      ]
    }
  },
  "top_p": 1
}

Example

curl https://api.openai.com/v1/assistants/asst_abc123 \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "OpenAI-Beta: assistants=v2"

Response

{
  "id": "asst_abc123",
  "object": "assistant",
  "created_at": 1699009709,
  "name": "HR Helper",
  "description": null,
  "model": "gpt-4o",
  "instructions": "You are an HR bot, and you have access to files to answer employee questions about company policies.",
  "tools": [
    {
      "type": "file_search"
    }
  ],
  "metadata": {},
  "top_p": 1.0,
  "temperature": 1.0,
  "response_format": "auto"
}

Modify assistant

post /assistants/{assistant_id}

Modifies an assistant.

Path Parameters

• assistant_id: string

Body Parameters

• description: optional string or null

  The description of the assistant. The maximum length is 512 characters.

• instructions: optional string or null

  The system instructions that the assistant uses. The maximum length is 256,000 characters.

• 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 format, and querying for objects via API or the dashboard.

  Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

• model: optional string or "gpt-5" or "gpt-5-mini" or "gpt-5-nano" or 39 more

  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

  • string

  • AssistantSupportedModels = "gpt-5" or "gpt-5-mini" or "gpt-5-nano" or 39 more

    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

    • "gpt-5"

    • "gpt-5-mini"

    • "gpt-5-nano"

    • "gpt-5-2025-08-07"

    • "gpt-5-mini-2025-08-07"

    • "gpt-5-nano-2025-08-07"

    • "gpt-4.1"

    • "gpt-4.1-mini"

    • "gpt-4.1-nano"

    • "gpt-4.1-2025-04-14"

    • "gpt-4.1-mini-2025-04-14"

    • "gpt-4.1-nano-2025-04-14"

    • "o3-mini"

    • "o3-mini-2025-01-31"

    • "o1"

    • "o1-2024-12-17"

    • "gpt-4o"

    • "gpt-4o-2024-11-20"

    • "gpt-4o-2024-08-06"

    • "gpt-4o-2024-05-13"

    • "gpt-4o-mini"

    • "gpt-4o-mini-2024-07-18"

    • "gpt-4.5-preview"

    • "gpt-4.5-preview-2025-02-27"

    • "gpt-4-turbo"

    • "gpt-4-turbo-2024-04-09"

    • "gpt-4-0125-preview"

    • "gpt-4-turbo-preview"

    • "gpt-4-1106-preview"

    • "gpt-4-vision-preview"

    • "gpt-4"

    • "gpt-4-0314"

    • "gpt-4-0613"

    • "gpt-4-32k"

    • "gpt-4-32k-0314"

    • "gpt-4-32k-0613"

    • "gpt-3.5-turbo"

    • "gpt-3.5-turbo-16k"

    • "gpt-3.5-turbo-0613"

    • "gpt-3.5-turbo-1106"

    • "gpt-3.5-turbo-0125"

    • "gpt-3.5-turbo-16k-0613"

• name: optional string or null

  The name of the assistant. The maximum length is 256 characters.

• reasoning_effort: optional ReasoningEffort or null

  Constrains effort on reasoning for reasoning models. Currently supported values are none, minimal, low, medium, high, xhigh, and max. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response. Not all reasoning models support every value. See the reasoning guide for model-specific support.

  • "none"

  • "minimal"

  • "low"

  • "medium"

  • "high"

  • "xhigh"

  • "max"

• response_format: optional AssistantResponseFormatOption or null

  Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

  Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

  Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

  Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

  • "auto"

    auto is the default value

    • "auto"

  • ResponseFormatText object { type }

    Default response format. Used to generate text responses.

    • type: "text"

      The type of response format being defined. Always text.

      • "text"

  • ResponseFormatJSONObject object { type }

    JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

    • type: "json_object"

      The type of response format being defined. Always json_object.

      • "json_object"

  • ResponseFormatJSONSchema object { json_schema, type }

    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

    • json_schema: object { name, description, schema, strict }

      Structured Outputs configuration options, including a JSON Schema.

      • name: string

        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

      • description: optional string

        A description of what the response format is for, used by the model to determine how to respond in the format.

      • schema: optional map[unknown]

        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

      • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

    • type: "json_schema"

      The type of response format being defined. Always json_schema.

      • "json_schema"

• 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.

• tool_resources: optional object { code_interpreter, file_search } or null

  A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

  • code_interpreter: optional object { file_ids }

    • file_ids: optional array of string

      Overrides the list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.

  • file_search: optional object { vector_store_ids }

    • vector_store_ids: optional array of string

      Overrides the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

• tools: optional array of CodeInterpreterTool or FileSearchTool or FunctionTool

  A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter, file_search, or function.

  • CodeInterpreterTool object { type }

    • type: "code_interpreter"

      The type of tool being defined: code_interpreter

      • "code_interpreter"

  • FileSearchTool object { type, file_search }

    • type: "file_search"

      The type of tool being defined: file_search

      • "file_search"

    • file_search: optional object { max_num_results, ranking_options }

      Overrides for the file search tool.

      • max_num_results: optional number

        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

      • ranking_options: optional object { score_threshold, ranker }

        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

        See the file search tool documentation for more information.

        • score_threshold: number

          The score threshold for the file search. All values must be a floating point number between 0 and 1.

        • ranker: optional "auto" or "default_2024_08_21"

          The ranker to use for the file search. If not specified will use the auto ranker.

          • "auto"

          • "default_2024_08_21"

  • FunctionTool object { function, type }

    • function: FunctionDefinition

      • name: string

        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

      • description: optional string

        A description of what the function does, used by the model to choose when and how to call the function.

      • parameters: optional FunctionParameters

        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

        Omitting parameters defines a function with an empty parameter list.

      • strict: optional boolean or null

        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

    • type: "function"

      The type of tool being defined: function

      • "function"

• 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 mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

  We generally recommend altering this or temperature but not both.

Returns

• Assistant object { id, created_at, description, 10 more }

  Represents an assistant that can call the model and use tools.

  • id: string

    The identifier, which can be referenced in API endpoints.

  • created_at: number

    The Unix timestamp (in seconds) for when the assistant was created.

  • description: string or null

    The description of the assistant. The maximum length is 512 characters.

  • instructions: string or null

    The system instructions that the assistant uses. The maximum length is 256,000 characters.

  • 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 format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • model: string

    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

  • name: string or null

    The name of the assistant. The maximum length is 256 characters.

  • object: "assistant"

    The object type, which is always assistant.

    • "assistant"

  • tools: array of CodeInterpreterTool or FileSearchTool or FunctionTool

    A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter, file_search, or function.

    • CodeInterpreterTool object { type }

      • type: "code_interpreter"

        The type of tool being defined: code_interpreter

        • "code_interpreter"

    • FileSearchTool object { type, file_search }

      • type: "file_search"

        The type of tool being defined: file_search

        • "file_search"

      • file_search: optional object { max_num_results, ranking_options }

        Overrides for the file search tool.

        • max_num_results: optional number

          The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

          Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

        • ranking_options: optional object { score_threshold, ranker }

          The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

          See the file search tool documentation for more information.

          • score_threshold: number

            The score threshold for the file search. All values must be a floating point number between 0 and 1.

          • ranker: optional "auto" or "default_2024_08_21"

            The ranker to use for the file search. If not specified will use the auto ranker.

            • "auto"

            • "default_2024_08_21"

    • FunctionTool object { function, type }

      • function: FunctionDefinition

        • name: string

          The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the function does, used by the model to choose when and how to call the function.

        • parameters: optional FunctionParameters

          The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

          Omitting parameters defines a function with an empty parameter list.

        • strict: optional boolean or null

          Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

      • type: "function"

        The type of tool being defined: function

        • "function"

  • response_format: optional AssistantResponseFormatOption or null

    Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

    Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

    Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

    Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

    • "auto"

      auto is the default value

      • "auto"

    • ResponseFormatText object { type }

      Default response format. Used to generate text responses.

      • type: "text"

        The type of response format being defined. Always text.

        • "text"

    • ResponseFormatJSONObject object { type }

      JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

      • type: "json_object"

        The type of response format being defined. Always json_object.

        • "json_object"

    • ResponseFormatJSONSchema object { json_schema, type }

      JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

      • json_schema: object { name, description, schema, strict }

        Structured Outputs configuration options, including a JSON Schema.

        • name: string

          The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the response format is for, used by the model to determine how to respond in the format.

        • schema: optional map[unknown]

          The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

        • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

      • type: "json_schema"

        The type of response format being defined. Always json_schema.

        • "json_schema"

  • 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.

  • tool_resources: optional object { code_interpreter, file_search } or null

    A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

    • code_interpreter: optional object { file_ids }

      • file_ids: optional array of string

        A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

    • file_search: optional object { vector_store_ids }

      • vector_store_ids: optional array of string

        The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

  • 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 mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    We generally recommend altering this or temperature but not both.

Example

curl https://api.openai.com/v1/assistants/$ASSISTANT_ID \
    -H 'Content-Type: application/json' \
    -H 'OpenAI-Beta: assistants=v2' \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d '{
          "temperature": 1,
          "top_p": 1
        }'

Response

{
  "id": "id",
  "created_at": 0,
  "description": "description",
  "instructions": "instructions",
  "metadata": {
    "foo": "string"
  },
  "model": "model",
  "name": "name",
  "object": "assistant",
  "tools": [
    {
      "type": "code_interpreter"
    }
  ],
  "response_format": "auto",
  "temperature": 1,
  "tool_resources": {
    "code_interpreter": {
      "file_ids": [
        "string"
      ]
    },
    "file_search": {
      "vector_store_ids": [
        "string"
      ]
    }
  },
  "top_p": 1
}

Example

curl https://api.openai.com/v1/assistants/asst_abc123 \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "OpenAI-Beta: assistants=v2" \
  -d '{
      "instructions": "You are an HR bot, and you have access to files to answer employee questions about company policies. Always response with info from either of the files.",
      "tools": [{"type": "file_search"}],
      "model": "gpt-4o"
    }'

Response

{
  "id": "asst_123",
  "object": "assistant",
  "created_at": 1699009709,
  "name": "HR Helper",
  "description": null,
  "model": "gpt-4o",
  "instructions": "You are an HR bot, and you have access to files to answer employee questions about company policies. Always response with info from either of the files.",
  "tools": [
    {
      "type": "file_search"
    }
  ],
  "tool_resources": {
    "file_search": {
      "vector_store_ids": []
    }
  },
  "metadata": {},
  "top_p": 1.0,
  "temperature": 1.0,
  "response_format": "auto"
}

Domain Types

Assistant

• Assistant object { id, created_at, description, 10 more }

  Represents an assistant that can call the model and use tools.

  • id: string

    The identifier, which can be referenced in API endpoints.

  • created_at: number

    The Unix timestamp (in seconds) for when the assistant was created.

  • description: string or null

    The description of the assistant. The maximum length is 512 characters.

  • instructions: string or null

    The system instructions that the assistant uses. The maximum length is 256,000 characters.

  • 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 format, and querying for objects via API or the dashboard.

    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

  • model: string

    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

  • name: string or null

    The name of the assistant. The maximum length is 256 characters.

  • object: "assistant"

    The object type, which is always assistant.

    • "assistant"

  • tools: array of CodeInterpreterTool or FileSearchTool or FunctionTool

    A list of tool enabled on the assistant. There can be a maximum of 128 tools per assistant. Tools can be of types code_interpreter, file_search, or function.

    • CodeInterpreterTool object { type }

      • type: "code_interpreter"

        The type of tool being defined: code_interpreter

        • "code_interpreter"

    • FileSearchTool object { type, file_search }

      • type: "file_search"

        The type of tool being defined: file_search

        • "file_search"

      • file_search: optional object { max_num_results, ranking_options }

        Overrides for the file search tool.

        • max_num_results: optional number

          The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

          Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

        • ranking_options: optional object { score_threshold, ranker }

          The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

          See the file search tool documentation for more information.

          • score_threshold: number

            The score threshold for the file search. All values must be a floating point number between 0 and 1.

          • ranker: optional "auto" or "default_2024_08_21"

            The ranker to use for the file search. If not specified will use the auto ranker.

            • "auto"

            • "default_2024_08_21"

    • FunctionTool object { function, type }

      • function: FunctionDefinition

        • name: string

          The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the function does, used by the model to choose when and how to call the function.

        • parameters: optional FunctionParameters

          The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

          Omitting parameters defines a function with an empty parameter list.

        • strict: optional boolean or null

          Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

      • type: "function"

        The type of tool being defined: function

        • "function"

  • response_format: optional AssistantResponseFormatOption or null

    Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

    Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

    Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

    Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

    • "auto"

      auto is the default value

      • "auto"

    • ResponseFormatText object { type }

      Default response format. Used to generate text responses.

      • type: "text"

        The type of response format being defined. Always text.

        • "text"

    • ResponseFormatJSONObject object { type }

      JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

      • type: "json_object"

        The type of response format being defined. Always json_object.

        • "json_object"

    • ResponseFormatJSONSchema object { json_schema, type }

      JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

      • json_schema: object { name, description, schema, strict }

        Structured Outputs configuration options, including a JSON Schema.

        • name: string

          The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

        • description: optional string

          A description of what the response format is for, used by the model to determine how to respond in the format.

        • schema: optional map[unknown]

          The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

        • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

      • type: "json_schema"

        The type of response format being defined. Always json_schema.

        • "json_schema"

  • 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.

  • tool_resources: optional object { code_interpreter, file_search } or null

    A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

    • code_interpreter: optional object { file_ids }

      • file_ids: optional array of string

        A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

    • file_search: optional object { vector_store_ids }

      • vector_store_ids: optional array of string

        The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

  • 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 mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.

    We generally recommend altering this or temperature but not both.

Assistant Deleted

• AssistantDeleted object { id, deleted, object }

  • id: string

  • deleted: boolean

  • object: "assistant.deleted"

    • "assistant.deleted"

Assistant Stream Event

• AssistantStreamEvent = object { data, event, enabled } or object { data, event } or object { data, event } or 22 more

  Represents an event emitted when streaming a Run.

  Each event in a server-sent events stream has an event and data property:

  event: thread.created
  data: {"id": "thread_123", "object": "thread", ...}
  

  We emit events whenever a new object is created, transitions to a new state, or is being streamed in parts (deltas). For example, we emit thread.run.created when a new run is created, thread.run.completed when a run completes, and so on. When an Assistant chooses to create a message during a run, we emit a thread.message.created event, a thread.message.in_progress event, many thread.message.delta events, and finally a thread.message.completed event.

  We may add additional events over time, so we recommend handling unknown events gracefully in your code. See the Assistants API quickstart to learn how to integrate the Assistants API with streaming.

  • ThreadCreated object { data, event, enabled }

    Occurs when a new thread is created.

    • data: Thread

      Represents a thread that contains messages.

      • id: string

        The identifier, which can be referenced in API endpoints.

      • created_at: number

        The Unix timestamp (in seconds) for when the thread was created.

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

      • object: "thread"

        The object type, which is always thread.

        • "thread"

      • tool_resources: object { code_interpreter, file_search } or null

        A set of resources that are made available to the assistant's tools in this thread. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

        • code_interpreter: optional object { file_ids }

          • file_ids: optional array of string

            A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.

        • file_search: optional object { vector_store_ids }

          • vector_store_ids: optional array of string

            The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.

    • event: "thread.created"

      • "thread.created"

    • enabled: optional boolean

      Whether to enable input audio transcription.

  • ThreadRunCreated object { data, event }

    Occurs when a new run is created.

    • data: Run

      Represents an execution run on a thread.

      • id: string

        The identifier, which can be referenced in API endpoints.

      • assistant_id: string

        The ID of the assistant used for execution of this run.

      • cancelled_at: number or null

        The Unix timestamp (in seconds) for when the run was cancelled.

      • completed_at: number or null

        The Unix timestamp (in seconds) for when the run was completed.

      • created_at: number

        The Unix timestamp (in seconds) for when the run was created.

      • expires_at: number or null

        The Unix timestamp (in seconds) for when the run will expire.

      • failed_at: number or null

        The Unix timestamp (in seconds) for when the run failed.

      • incomplete_details: object { reason } or null

        Details on why the run is incomplete. Will be null if the run is not incomplete.

        • reason: optional "max_completion_tokens" or "max_prompt_tokens"

          The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.

          • "max_completion_tokens"

          • "max_prompt_tokens"

      • instructions: string

        The instructions that the assistant used for this run.

      • last_error: object { code, message } or null

        The last error associated with this run. Will be null if there are no errors.

        • code: "server_error" or "rate_limit_exceeded" or "invalid_prompt"

          One of server_error, rate_limit_exceeded, or invalid_prompt.

          • "server_error"

          • "rate_limit_exceeded"

          • "invalid_prompt"

        • message: string

          A human-readable description of the error.

      • max_completion_tokens: number or null

        The maximum number of completion tokens specified to have been used over the course of the run.

      • max_prompt_tokens: number or null

        The maximum number of prompt tokens specified to have been used over the course of the run.

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

      • model: string

        The model that the assistant used for this run.

      • object: "thread.run"

        The object type, which is always thread.run.

        • "thread.run"

      • parallel_tool_calls: boolean

        Whether to enable parallel function calling during tool use.

      • required_action: object { submit_tool_outputs, type } or null

        Details on the action required to continue the run. Will be null if no action is required.

        • submit_tool_outputs: object { tool_calls }

          Details on the tool outputs needed for this run to continue.

          • tool_calls: array of RequiredActionFunctionToolCall

            A list of the relevant tool calls.

            • id: string

              The ID of the tool call. This ID must be referenced when you submit the tool outputs in using the Submit tool outputs to run endpoint.

            • function: object { arguments, name }

              The function definition.

              • arguments: string

                The arguments that the model expects you to pass to the function.

              • name: string

                The name of the function.

            • type: "function"

              The type of tool call the output is required for. For now, this is always function.

              • "function"

        • type: "submit_tool_outputs"

          For now, this is always submit_tool_outputs.

          • "submit_tool_outputs"

      • response_format: AssistantResponseFormatOption or null

        Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

        Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

        Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

        Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

        • "auto"

          auto is the default value

          • "auto"

        • ResponseFormatText object { type }

          Default response format. Used to generate text responses.

          • type: "text"

            The type of response format being defined. Always text.

            • "text"

        • ResponseFormatJSONObject object { type }

          JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

          • type: "json_object"

            The type of response format being defined. Always json_object.

            • "json_object"

        • ResponseFormatJSONSchema object { json_schema, type }

          JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

          • json_schema: object { name, description, schema, strict }

            Structured Outputs configuration options, including a JSON Schema.

            • name: string

              The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

            • description: optional string

              A description of what the response format is for, used by the model to determine how to respond in the format.

            • schema: optional map[unknown]

              The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

            • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

          • type: "json_schema"

            The type of response format being defined. Always json_schema.

            • "json_schema"

      • started_at: number or null

        The Unix timestamp (in seconds) for when the run was started.

      • status: "queued" or "in_progress" or "requires_action" or 6 more

        The status of the run, which can be either queued, in_progress, requires_action, cancelling, cancelled, failed, completed, incomplete, or expired.

        • "queued"

        • "in_progress"

        • "requires_action"

        • "cancelling"

        • "cancelled"

        • "failed"

        • "completed"

        • "incomplete"

        • "expired"

      • thread_id: string

        The ID of the thread that was executed on as a part of this run.

      • tool_choice: AssistantToolChoiceOption or null

        Controls which (if any) tool is called by the model. none means the model will not call any tools and instead generates a message. auto is the default value and means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools before responding to the user. Specifying a particular tool like {"type": "file_search"} or {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool.

        • "none" or "auto" or "required"

          none means the model will not call any tools and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools before responding to the user.

          • "none"

          • "auto"

          • "required"

        • AssistantToolChoice object { type, function }

          Specifies a tool the model should use. Use to force the model to call a specific tool.

          • type: "function" or "code_interpreter" or "file_search"

            The type of the tool. If type is function, the function name must be set

            • "function"

            • "code_interpreter"

            • "file_search"

          • function: optional AssistantToolChoiceFunction

            • name: string

              The name of the function to call.

      • tools: array of CodeInterpreterTool or FileSearchTool or FunctionTool

        The list of tools that the assistant used for this run.

        • CodeInterpreterTool object { type }

          • type: "code_interpreter"

            The type of tool being defined: code_interpreter

            • "code_interpreter"

        • FileSearchTool object { type, file_search }

          • type: "file_search"

            The type of tool being defined: file_search

            • "file_search"

          • file_search: optional object { max_num_results, ranking_options }

            Overrides for the file search tool.

            • max_num_results: optional number

              The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

              Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

            • ranking_options: optional object { score_threshold, ranker }

              The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

              See the file search tool documentation for more information.

              • score_threshold: number

                The score threshold for the file search. All values must be a floating point number between 0 and 1.

              • ranker: optional "auto" or "default_2024_08_21"

                The ranker to use for the file search. If not specified will use the auto ranker.

                • "auto"

                • "default_2024_08_21"

        • FunctionTool object { function, type }

          • function: FunctionDefinition

            • name: string

              The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

            • description: optional string

              A description of what the function does, used by the model to choose when and how to call the function.

            • parameters: optional FunctionParameters

              The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

              Omitting parameters defines a function with an empty parameter list.

            • strict: optional boolean or null

              Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

          • type: "function"

            The type of tool being defined: function

            • "function"

      • truncation_strategy: object { type, last_messages } or null

        Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

        • type: "auto" or "last_messages"

          The truncation strategy to use for the thread. The default is auto. If set to last_messages, the thread will be truncated to the n most recent messages in the thread. When set to auto, messages in the middle of the thread will be dropped to fit the context length of the model, max_prompt_tokens.

          • "auto"

          • "last_messages"

        • last_messages: optional number or null

          The number of most recent messages from the thread when constructing the context for the run.

      • usage: object { completion_tokens, prompt_tokens, total_tokens } or null

        Usage statistics related to the run. This value will be null if the run is not in a terminal state (i.e. in_progress, queued, etc.).

        • completion_tokens: number

          Number of completion tokens used over the course of the run.

        • prompt_tokens: number

          Number of prompt tokens used over the course of the run.

        • total_tokens: number

          Total number of tokens used (prompt + completion).

      • temperature: optional number or null

        The sampling temperature used for this run. If not set, defaults to 1.

      • top_p: optional number or null

        The nucleus sampling value used for this run. If not set, defaults to 1.

    • event: "thread.run.created"

      • "thread.run.created"

  • ThreadRunQueued object { data, event }

    Occurs when a run moves to a queued status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.queued"

      • "thread.run.queued"

  • ThreadRunInProgress object { data, event }

    Occurs when a run moves to an in_progress status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.in_progress"

      • "thread.run.in_progress"

  • ThreadRunRequiresAction object { data, event }

    Occurs when a run moves to a requires_action status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.requires_action"

      • "thread.run.requires_action"

  • ThreadRunCompleted object { data, event }

    Occurs when a run is completed.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.completed"

      • "thread.run.completed"

  • ThreadRunIncomplete object { data, event }

    Occurs when a run ends with status incomplete.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.incomplete"

      • "thread.run.incomplete"

  • ThreadRunFailed object { data, event }

    Occurs when a run fails.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.failed"

      • "thread.run.failed"

  • ThreadRunCancelling object { data, event }

    Occurs when a run moves to a cancelling status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.cancelling"

      • "thread.run.cancelling"

  • ThreadRunCancelled object { data, event }

    Occurs when a run is cancelled.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.cancelled"

      • "thread.run.cancelled"

  • ThreadRunExpired object { data, event }

    Occurs when a run expires.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.expired"

      • "thread.run.expired"

  • ThreadRunStepCreated object { data, event }

    Occurs when a run step is created.

    • data: RunStep

      Represents a step in execution of a run.

      • id: string

        The identifier of the run step, which can be referenced in API endpoints.

      • assistant_id: string

        The ID of the assistant associated with the run step.

      • cancelled_at: number or null

        The Unix timestamp (in seconds) for when the run step was cancelled.

      • completed_at: number or null

        The Unix timestamp (in seconds) for when the run step completed.

      • created_at: number

        The Unix timestamp (in seconds) for when the run step was created.

      • expired_at: number or null

        The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.

      • failed_at: number or null

        The Unix timestamp (in seconds) for when the run step failed.

      • last_error: object { code, message } or null

        The last error associated with this run step. Will be null if there are no errors.

        • code: "server_error" or "rate_limit_exceeded"

          One of server_error or rate_limit_exceeded.

          • "server_error"

          • "rate_limit_exceeded"

        • message: string

          A human-readable description of the error.

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

      • object: "thread.run.step"

        The object type, which is always thread.run.step.

        • "thread.run.step"

      • run_id: string

        The ID of the run that this run step is a part of.

      • status: "in_progress" or "cancelled" or "failed" or 2 more

        The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.

        • "in_progress"

        • "cancelled"

        • "failed"

        • "completed"

        • "expired"

      • step_details: MessageCreationStepDetails or ToolCallsStepDetails

        The details of the run step.

        • MessageCreationStepDetails object { message_creation, type }

          Details of the message creation by the run step.

          • message_creation: object { message_id }

            • message_id: string

              The ID of the message that was created by this run step.

          • type: "message_creation"

            Always message_creation.

            • "message_creation"

        • ToolCallsStepDetails object { tool_calls, type }

          Details of the tool call.

          • tool_calls: array of CodeInterpreterToolCall or FileSearchToolCall or FunctionToolCall

            An array of tool calls the run step was involved in. These can be associated with one of three types of tools: code_interpreter, file_search, or function.

            • CodeInterpreterToolCall object { id, code_interpreter, type }

              Details of the Code Interpreter tool call the run step was involved in.

              • id: string

                The ID of the tool call.

              • code_interpreter: object { input, outputs }

                The Code Interpreter tool call definition.

                • input: string

                  The input to the Code Interpreter tool call.

                • outputs: array of object { logs, type } or object { image, type }

                  The outputs from the Code Interpreter tool call. Code Interpreter can output one or more items, including text (logs) or images (image). Each of these are represented by a different object type.

                  • CodeInterpreterLogOutput object { logs, type }

                    Text output from the Code Interpreter tool call as part of a run step.

                    • logs: string

                      The text output from the Code Interpreter tool call.

                    • type: "logs"

                      Always logs.

                      • "logs"

                  • CodeInterpreterImageOutput object { image, type }

                    • image: object { file_id }

                      • file_id: string

                        The file ID of the image.

                    • type: "image"

                      Always image.

                      • "image"

              • type: "code_interpreter"

                The type of tool call. This is always going to be code_interpreter for this type of tool call.

                • "code_interpreter"

            • FileSearchToolCall object { id, file_search, type }

              • id: string

                The ID of the tool call object.

              • file_search: object { ranking_options, results }

                For now, this is always going to be an empty object.

                • ranking_options: optional object { ranker, score_threshold }

                  The ranking options for the file search.

                  • ranker: "auto" or "default_2024_08_21"

                    The ranker to use for the file search. If not specified will use the auto ranker.

                    • "auto"

                    • "default_2024_08_21"

                  • score_threshold: number

                    The score threshold for the file search. All values must be a floating point number between 0 and 1.

                • results: optional array of object { file_id, file_name, score, content }

                  The results of the file search.

                  • file_id: string

                    The ID of the file that result was found in.

                  • file_name: string

                    The name of the file that result was found in.

                  • score: number

                    The score of the result. All values must be a floating point number between 0 and 1.

                  • content: optional array of object { text, type }

                    The content of the result that was found. The content is only included if requested via the include query parameter.

                    • text: optional string

                      The text content of the file.

                    • type: optional "text"

                      The type of the content.

                      • "text"

              • type: "file_search"

                The type of tool call. This is always going to be file_search for this type of tool call.

                • "file_search"

            • FunctionToolCall object { id, function, type }

              • id: string

                The ID of the tool call object.

              • function: object { arguments, name, output }

                The definition of the function that was called.

                • arguments: string

                  The arguments passed to the function.

                • name: string

                  The name of the function.

                • output: string or null

                  The output of the function. This will be null if the outputs have not been submitted yet.

              • type: "function"

                The type of tool call. This is always going to be function for this type of tool call.

                • "function"

          • type: "tool_calls"

            Always tool_calls.

            • "tool_calls"

      • thread_id: string

        The ID of the thread that was run.

      • type: "message_creation" or "tool_calls"

        The type of run step, which can be either message_creation or tool_calls.

        • "message_creation"

        • "tool_calls"

      • usage: object { completion_tokens, prompt_tokens, total_tokens } or null

        Usage statistics related to the run step. This value will be null while the run step's status is in_progress.

        • completion_tokens: number

          Number of completion tokens used over the course of the run step.

        • prompt_tokens: number

          Number of prompt tokens used over the course of the run step.

        • total_tokens: number

          Total number of tokens used (prompt + completion).

    • event: "thread.run.step.created"

      • "thread.run.step.created"

  • ThreadRunStepInProgress object { data, event }

    Occurs when a run step moves to an in_progress state.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.in_progress"

      • "thread.run.step.in_progress"

  • ThreadRunStepDelta object { data, event }

    Occurs when parts of a run step are being streamed.

    • data: RunStepDeltaEvent

      Represents a run step delta i.e. any changed fields on a run step during streaming.

      • id: string

        The identifier of the run step, which can be referenced in API endpoints.

      • delta: object { step_details }

        The delta containing the fields that have changed on the run step.

        • step_details: optional RunStepDeltaMessageDelta or ToolCallDeltaObject

          The details of the run step.

          • RunStepDeltaMessageDelta object { type, message_creation }

            Details of the message creation by the run step.

            • type: "message_creation"

              Always message_creation.

              • "message_creation"

            • message_creation: optional object { message_id }

              • message_id: optional string

                The ID of the message that was created by this run step.

          • ToolCallDeltaObject object { type, tool_calls }

            Details of the tool call.

            • type: "tool_calls"

              Always tool_calls.

              • "tool_calls"

            • tool_calls: optional array of CodeInterpreterToolCallDelta or FileSearchToolCallDelta or FunctionToolCallDelta

              An array of tool calls the run step was involved in. These can be associated with one of three types of tools: code_interpreter, file_search, or function.

              • CodeInterpreterToolCallDelta object { index, type, id, code_interpreter }

                Details of the Code Interpreter tool call the run step was involved in.

                • index: number

                  The index of the tool call in the tool calls array.

                • type: "code_interpreter"

                  The type of tool call. This is always going to be code_interpreter for this type of tool call.

                  • "code_interpreter"

                • id: optional string

                  The ID of the tool call.

                • code_interpreter: optional object { input, outputs }

                  The Code Interpreter tool call definition.

                  • input: optional string

                    The input to the Code Interpreter tool call.

                  • outputs: optional array of CodeInterpreterLogs or CodeInterpreterOutputImage

                    The outputs from the Code Interpreter tool call. Code Interpreter can output one or more items, including text (logs) or images (image). Each of these are represented by a different object type.

                    • CodeInterpreterLogs object { index, type, logs }

                      Text output from the Code Interpreter tool call as part of a run step.

                      • index: number

                        The index of the output in the outputs array.

                      • type: "logs"

                        Always logs.

                        • "logs"

                      • logs: optional string

                        The text output from the Code Interpreter tool call.

                    • CodeInterpreterOutputImage object { index, type, image }

                      • index: number

                        The index of the output in the outputs array.

                      • type: "image"

                        Always image.

                        • "image"

                      • image: optional object { file_id }

                        • file_id: optional string

                          The file ID of the image.

              • FileSearchToolCallDelta object { file_search, index, type, id }

                • file_search: unknown

                  For now, this is always going to be an empty object.

                • index: number

                  The index of the tool call in the tool calls array.

                • type: "file_search"

                  The type of tool call. This is always going to be file_search for this type of tool call.

                  • "file_search"

                • id: optional string

                  The ID of the tool call object.

              • FunctionToolCallDelta object { index, type, id, function }

                • index: number

                  The index of the tool call in the tool calls array.

                • type: "function"

                  The type of tool call. This is always going to be function for this type of tool call.

                  • "function"

                • id: optional string

                  The ID of the tool call object.

                • function: optional object { arguments, name, output }

                  The definition of the function that was called.

                  • arguments: optional string

                    The arguments passed to the function.

                  • name: optional string

                    The name of the function.

                  • output: optional string or null

                    The output of the function. This will be null if the outputs have not been submitted yet.

      • object: "thread.run.step.delta"

        The object type, which is always thread.run.step.delta.

        • "thread.run.step.delta"

    • event: "thread.run.step.delta"

      • "thread.run.step.delta"

  • ThreadRunStepCompleted object { data, event }

    Occurs when a run step is completed.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.completed"

      • "thread.run.step.completed"

  • ThreadRunStepFailed object { data, event }

    Occurs when a run step fails.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.failed"

      • "thread.run.step.failed"

  • ThreadRunStepCancelled object { data, event }

    Occurs when a run step is cancelled.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.cancelled"

      • "thread.run.step.cancelled"

  • ThreadRunStepExpired object { data, event }

    Occurs when a run step expires.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.expired"

      • "thread.run.step.expired"

  • ThreadMessageCreated object { data, event }

    Occurs when a message is created.

    • data: Message

      Represents a message within a thread.

      • id: string

        The identifier, which can be referenced in API endpoints.

      • assistant_id: string or null

        If applicable, the ID of the assistant that authored this message.

      • attachments: array of object { file_id, tools } or null

        A list of files attached to the message, and the tools they were added to.

        • file_id: optional string

          The ID of the file to attach to the message.

        • tools: optional array of CodeInterpreterTool or object { type }

          The tools to add this file to.

          • CodeInterpreterTool object { type }

          • FileSearchTool object { type }

            • type: "file_search"

              The type of tool being defined: file_search

              • "file_search"

      • completed_at: number or null

        The Unix timestamp (in seconds) for when the message was completed.

      • content: array of ImageFileContentBlock or ImageURLContentBlock or TextContentBlock or RefusalContentBlock

        The content of the message in array of text and/or images.

        • ImageFileContentBlock object { image_file, type }

          References an image File in the content of a message.

          • image_file: ImageFile

            • file_id: string

              The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.

            • detail: optional "auto" or "low" or "high"

              Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.

              • "auto"

              • "low"

              • "high"

          • type: "image_file"

            Always image_file.

            • "image_file"

        • ImageURLContentBlock object { image_url, type }

          References an image URL in the content of a message.

          • image_url: ImageURL

            • url: string

              The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

            • detail: optional "auto" or "low" or "high"

              Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high. Default value is auto

              • "auto"

              • "low"

              • "high"

          • type: "image_url"

            The type of the content part.

            • "image_url"

        • TextContentBlock object { text, type }

          The text content that is part of a message.

          • text: Text

            • annotations: array of FileCitationAnnotation or FilePathAnnotation

              • FileCitationAnnotation object { end_index, file_citation, start_index, 2 more }

                A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.

                • end_index: number

                • file_citation: object { file_id }

                  • file_id: string

                    The ID of the specific File the citation is from.

                • start_index: number

                • text: string

                  The text in the message content that needs to be replaced.

                • type: "file_citation"

                  Always file_citation.

                  • "file_citation"

              • FilePathAnnotation object { end_index, file_path, start_index, 2 more }

                A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.

                • end_index: number

                • file_path: object { file_id }

                  • file_id: string

                    The ID of the file that was generated.

                • start_index: number

                • text: string

                  The text in the message content that needs to be replaced.

                • type: "file_path"

                  Always file_path.

                  • "file_path"

            • value: string

              The data that makes up the text.

          • type: "text"

            Always text.

            • "text"

        • RefusalContentBlock object { refusal, type }

          The refusal content generated by the assistant.

          • refusal: string

          • type: "refusal"

            Always refusal.

            • "refusal"

      • created_at: number

        The Unix timestamp (in seconds) for when the message was created.

      • incomplete_at: number or null

        The Unix timestamp (in seconds) for when the message was marked as incomplete.

      • incomplete_details: object { reason } or null

        On an incomplete message, details about why the message is incomplete.

        • reason: "content_filter" or "max_tokens" or "run_cancelled" or 2 more

          The reason the message is incomplete.

          • "content_filter"

          • "max_tokens"

          • "run_cancelled"

          • "run_expired"

          • "run_failed"

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

      • object: "thread.message"

        The object type, which is always thread.message.

        • "thread.message"

      • role: "user" or "assistant"

        The entity that produced the message. One of user or assistant.

        • "user"

        • "assistant"

      • run_id: string or null

        The ID of the run associated with the creation of this message. Value is null when messages are created manually using the create message or create thread endpoints.

      • status: "in_progress" or "incomplete" or "completed"

        The status of the message, which can be either in_progress, incomplete, or completed.

        • "in_progress"

        • "incomplete"

        • "completed"

      • thread_id: string

        The thread ID that this message belongs to.

    • event: "thread.message.created"

      • "thread.message.created"

  • ThreadMessageInProgress object { data, event }

    Occurs when a message moves to an in_progress state.

    • data: Message

      Represents a message within a thread.

    • event: "thread.message.in_progress"

      • "thread.message.in_progress"

  • ThreadMessageDelta object { data, event }

    Occurs when parts of a Message are being streamed.

    • data: MessageDeltaEvent

      Represents a message delta i.e. any changed fields on a message during streaming.

      • id: string

        The identifier of the message, which can be referenced in API endpoints.

      • delta: MessageDelta

        The delta containing the fields that have changed on the Message.

        • content: optional array of ImageFileDeltaBlock or TextDeltaBlock or RefusalDeltaBlock or ImageURLDeltaBlock

          The content of the message in array of text and/or images.

          • ImageFileDeltaBlock object { index, type, image_file }

            References an image File in the content of a message.

            • index: number

              The index of the content part in the message.

            • type: "image_file"

              Always image_file.

              • "image_file"

            • image_file: optional ImageFileDelta

              • detail: optional "auto" or "low" or "high"

                Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.

                • "auto"

                • "low"

                • "high"

              • file_id: optional string

                The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.

          • TextDeltaBlock object { index, type, text }

            The text content that is part of a message.

            • index: number

              The index of the content part in the message.

            • type: "text"

              Always text.

              • "text"

            • text: optional TextDelta

              • annotations: optional array of FileCitationDeltaAnnotation or FilePathDeltaAnnotation

                • FileCitationDeltaAnnotation object { index, type, end_index, 3 more }

                  A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.

                  • index: number

                    The index of the annotation in the text content part.

                  • type: "file_citation"

                    Always file_citation.

                    • "file_citation"

                  • end_index: optional number

                  • file_citation: optional object { file_id, quote }

                    • file_id: optional string

                      The ID of the specific File the citation is from.

                    • quote: optional string

                      The specific quote in the file.

                  • start_index: optional number

                  • text: optional string

                    The text in the message content that needs to be replaced.

                • FilePathDeltaAnnotation object { index, type, end_index, 3 more }

                  A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.

                  • index: number

                    The index of the annotation in the text content part.

                  • type: "file_path"

                    Always file_path.

                    • "file_path"

                  • end_index: optional number

                  • file_path: optional object { file_id }

                    • file_id: optional string

                      The ID of the file that was generated.

                  • start_index: optional number

                  • text: optional string

                    The text in the message content that needs to be replaced.

              • value: optional string

                The data that makes up the text.

          • RefusalDeltaBlock object { index, type, refusal }

            The refusal content that is part of a message.

            • index: number

              The index of the refusal part in the message.

            • type: "refusal"

              Always refusal.

              • "refusal"

            • refusal: optional string

          • ImageURLDeltaBlock object { index, type, image_url }

            References an image URL in the content of a message.

            • index: number

              The index of the content part in the message.

            • type: "image_url"

              Always image_url.

              • "image_url"

            • image_url: optional ImageURLDelta

              • detail: optional "auto" or "low" or "high"

                Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high.

                • "auto"

                • "low"

                • "high"

              • url: optional string

                The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

        • role: optional "user" or "assistant"

          The entity that produced the message. One of user or assistant.

          • "user"

          • "assistant"

      • object: "thread.message.delta"

        The object type, which is always thread.message.delta.

        • "thread.message.delta"

    • event: "thread.message.delta"

      • "thread.message.delta"

  • ThreadMessageCompleted object { data, event }

    Occurs when a message is completed.

    • data: Message

      Represents a message within a thread.

    • event: "thread.message.completed"

      • "thread.message.completed"

  • ThreadMessageIncomplete object { data, event }

    Occurs when a message ends before it is completed.

    • data: Message

      Represents a message within a thread.

    • event: "thread.message.incomplete"

      • "thread.message.incomplete"

  • ErrorEvent object { data, event }

    Occurs when an error occurs. This can happen due to an internal server error or a timeout.

    • data: ErrorObject

      • code: string or null

      • message: string

      • param: string or null

      • type: string

    • event: "error"

      • "error"

  • DoneEvent object { data, event }

    Occurs when a stream ends.

    • data: "[DONE]"

      • "[DONE]"

    • event: "done"

      • "done"

Code Interpreter Tool

• CodeInterpreterTool object { type }

  • type: "code_interpreter"

    The type of tool being defined: code_interpreter

    • "code_interpreter"

File Search Tool

• FileSearchTool object { type, file_search }

  • type: "file_search"

    The type of tool being defined: file_search

    • "file_search"

  • file_search: optional object { max_num_results, ranking_options }

    Overrides for the file search tool.

    • max_num_results: optional number

      The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

      Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

    • ranking_options: optional object { score_threshold, ranker }

      The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

      See the file search tool documentation for more information.

      • score_threshold: number

        The score threshold for the file search. All values must be a floating point number between 0 and 1.

      • ranker: optional "auto" or "default_2024_08_21"

        The ranker to use for the file search. If not specified will use the auto ranker.

        • "auto"

        • "default_2024_08_21"

Function Tool

• FunctionTool object { function, type }

  • function: FunctionDefinition

    • name: string

      The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

    • description: optional string

      A description of what the function does, used by the model to choose when and how to call the function.

    • parameters: optional FunctionParameters

      The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

      Omitting parameters defines a function with an empty parameter list.

    • strict: optional boolean or null

      Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

  • type: "function"

    The type of tool being defined: function

    • "function"

Message Stream Event

• MessageStreamEvent = object { data, event } or object { data, event } or object { data, event } or 2 more

  Occurs when a message is created.

  • ThreadMessageCreated object { data, event }

    Occurs when a message is created.

    • data: Message

      Represents a message within a thread.

      • id: string

        The identifier, which can be referenced in API endpoints.

      • assistant_id: string or null

        If applicable, the ID of the assistant that authored this message.

      • attachments: array of object { file_id, tools } or null

        A list of files attached to the message, and the tools they were added to.

        • file_id: optional string

          The ID of the file to attach to the message.

        • tools: optional array of CodeInterpreterTool or object { type }

          The tools to add this file to.

          • CodeInterpreterTool object { type }

            • type: "code_interpreter"

              The type of tool being defined: code_interpreter

              • "code_interpreter"

          • FileSearchTool object { type }

            • type: "file_search"

              The type of tool being defined: file_search

              • "file_search"

      • completed_at: number or null

        The Unix timestamp (in seconds) for when the message was completed.

      • content: array of ImageFileContentBlock or ImageURLContentBlock or TextContentBlock or RefusalContentBlock

        The content of the message in array of text and/or images.

        • ImageFileContentBlock object { image_file, type }

          References an image File in the content of a message.

          • image_file: ImageFile

            • file_id: string

              The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.

            • detail: optional "auto" or "low" or "high"

              Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.

              • "auto"

              • "low"

              • "high"

          • type: "image_file"

            Always image_file.

            • "image_file"

        • ImageURLContentBlock object { image_url, type }

          References an image URL in the content of a message.

          • image_url: ImageURL

            • url: string

              The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

            • detail: optional "auto" or "low" or "high"

              Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high. Default value is auto

              • "auto"

              • "low"

              • "high"

          • type: "image_url"

            The type of the content part.

            • "image_url"

        • TextContentBlock object { text, type }

          The text content that is part of a message.

          • text: Text

            • annotations: array of FileCitationAnnotation or FilePathAnnotation

              • FileCitationAnnotation object { end_index, file_citation, start_index, 2 more }

                A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.

                • end_index: number

                • file_citation: object { file_id }

                  • file_id: string

                    The ID of the specific File the citation is from.

                • start_index: number

                • text: string

                  The text in the message content that needs to be replaced.

                • type: "file_citation"

                  Always file_citation.

                  • "file_citation"

              • FilePathAnnotation object { end_index, file_path, start_index, 2 more }

                A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.

                • end_index: number

                • file_path: object { file_id }

                  • file_id: string

                    The ID of the file that was generated.

                • start_index: number

                • text: string

                  The text in the message content that needs to be replaced.

                • type: "file_path"

                  Always file_path.

                  • "file_path"

            • value: string

              The data that makes up the text.

          • type: "text"

            Always text.

            • "text"

        • RefusalContentBlock object { refusal, type }

          The refusal content generated by the assistant.

          • refusal: string

          • type: "refusal"

            Always refusal.

            • "refusal"

      • created_at: number

        The Unix timestamp (in seconds) for when the message was created.

      • incomplete_at: number or null

        The Unix timestamp (in seconds) for when the message was marked as incomplete.

      • incomplete_details: object { reason } or null

        On an incomplete message, details about why the message is incomplete.

        • reason: "content_filter" or "max_tokens" or "run_cancelled" or 2 more

          The reason the message is incomplete.

          • "content_filter"

          • "max_tokens"

          • "run_cancelled"

          • "run_expired"

          • "run_failed"

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

      • object: "thread.message"

        The object type, which is always thread.message.

        • "thread.message"

      • role: "user" or "assistant"

        The entity that produced the message. One of user or assistant.

        • "user"

        • "assistant"

      • run_id: string or null

        The ID of the run associated with the creation of this message. Value is null when messages are created manually using the create message or create thread endpoints.

      • status: "in_progress" or "incomplete" or "completed"

        The status of the message, which can be either in_progress, incomplete, or completed.

        • "in_progress"

        • "incomplete"

        • "completed"

      • thread_id: string

        The thread ID that this message belongs to.

    • event: "thread.message.created"

      • "thread.message.created"

  • ThreadMessageInProgress object { data, event }

    Occurs when a message moves to an in_progress state.

    • data: Message

      Represents a message within a thread.

    • event: "thread.message.in_progress"

      • "thread.message.in_progress"

  • ThreadMessageDelta object { data, event }

    Occurs when parts of a Message are being streamed.

    • data: MessageDeltaEvent

      Represents a message delta i.e. any changed fields on a message during streaming.

      • id: string

        The identifier of the message, which can be referenced in API endpoints.

      • delta: MessageDelta

        The delta containing the fields that have changed on the Message.

        • content: optional array of ImageFileDeltaBlock or TextDeltaBlock or RefusalDeltaBlock or ImageURLDeltaBlock

          The content of the message in array of text and/or images.

          • ImageFileDeltaBlock object { index, type, image_file }

            References an image File in the content of a message.

            • index: number

              The index of the content part in the message.

            • type: "image_file"

              Always image_file.

              • "image_file"

            • image_file: optional ImageFileDelta

              • detail: optional "auto" or "low" or "high"

                Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.

                • "auto"

                • "low"

                • "high"

              • file_id: optional string

                The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.

          • TextDeltaBlock object { index, type, text }

            The text content that is part of a message.

            • index: number

              The index of the content part in the message.

            • type: "text"

              Always text.

              • "text"

            • text: optional TextDelta

              • annotations: optional array of FileCitationDeltaAnnotation or FilePathDeltaAnnotation

                • FileCitationDeltaAnnotation object { index, type, end_index, 3 more }

                  A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.

                  • index: number

                    The index of the annotation in the text content part.

                  • type: "file_citation"

                    Always file_citation.

                    • "file_citation"

                  • end_index: optional number

                  • file_citation: optional object { file_id, quote }

                    • file_id: optional string

                      The ID of the specific File the citation is from.

                    • quote: optional string

                      The specific quote in the file.

                  • start_index: optional number

                  • text: optional string

                    The text in the message content that needs to be replaced.

                • FilePathDeltaAnnotation object { index, type, end_index, 3 more }

                  A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.

                  • index: number

                    The index of the annotation in the text content part.

                  • type: "file_path"

                    Always file_path.

                    • "file_path"

                  • end_index: optional number

                  • file_path: optional object { file_id }

                    • file_id: optional string

                      The ID of the file that was generated.

                  • start_index: optional number

                  • text: optional string

                    The text in the message content that needs to be replaced.

              • value: optional string

                The data that makes up the text.

          • RefusalDeltaBlock object { index, type, refusal }

            The refusal content that is part of a message.

            • index: number

              The index of the refusal part in the message.

            • type: "refusal"

              Always refusal.

              • "refusal"

            • refusal: optional string

          • ImageURLDeltaBlock object { index, type, image_url }

            References an image URL in the content of a message.

            • index: number

              The index of the content part in the message.

            • type: "image_url"

              Always image_url.

              • "image_url"

            • image_url: optional ImageURLDelta

              • detail: optional "auto" or "low" or "high"

                Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high.

                • "auto"

                • "low"

                • "high"

              • url: optional string

                The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

        • role: optional "user" or "assistant"

          The entity that produced the message. One of user or assistant.

          • "user"

          • "assistant"

      • object: "thread.message.delta"

        The object type, which is always thread.message.delta.

        • "thread.message.delta"

    • event: "thread.message.delta"

      • "thread.message.delta"

  • ThreadMessageCompleted object { data, event }

    Occurs when a message is completed.

    • data: Message

      Represents a message within a thread.

    • event: "thread.message.completed"

      • "thread.message.completed"

  • ThreadMessageIncomplete object { data, event }

    Occurs when a message ends before it is completed.

    • data: Message

      Represents a message within a thread.

    • event: "thread.message.incomplete"

      • "thread.message.incomplete"

Run Step Stream Event

• RunStepStreamEvent = object { data, event } or object { data, event } or object { data, event } or 4 more

  Occurs when a run step is created.

  • ThreadRunStepCreated object { data, event }

    Occurs when a run step is created.

    • data: RunStep

      Represents a step in execution of a run.

      • id: string

        The identifier of the run step, which can be referenced in API endpoints.

      • assistant_id: string

        The ID of the assistant associated with the run step.

      • cancelled_at: number or null

        The Unix timestamp (in seconds) for when the run step was cancelled.

      • completed_at: number or null

        The Unix timestamp (in seconds) for when the run step completed.

      • created_at: number

        The Unix timestamp (in seconds) for when the run step was created.

      • expired_at: number or null

        The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.

      • failed_at: number or null

        The Unix timestamp (in seconds) for when the run step failed.

      • last_error: object { code, message } or null

        The last error associated with this run step. Will be null if there are no errors.

        • code: "server_error" or "rate_limit_exceeded"

          One of server_error or rate_limit_exceeded.

          • "server_error"

          • "rate_limit_exceeded"

        • message: string

          A human-readable description of the error.

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

      • object: "thread.run.step"

        The object type, which is always thread.run.step.

        • "thread.run.step"

      • run_id: string

        The ID of the run that this run step is a part of.

      • status: "in_progress" or "cancelled" or "failed" or 2 more

        The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.

        • "in_progress"

        • "cancelled"

        • "failed"

        • "completed"

        • "expired"

      • step_details: MessageCreationStepDetails or ToolCallsStepDetails

        The details of the run step.

        • MessageCreationStepDetails object { message_creation, type }

          Details of the message creation by the run step.

          • message_creation: object { message_id }

            • message_id: string

              The ID of the message that was created by this run step.

          • type: "message_creation"

            Always message_creation.

            • "message_creation"

        • ToolCallsStepDetails object { tool_calls, type }

          Details of the tool call.

          • tool_calls: array of CodeInterpreterToolCall or FileSearchToolCall or FunctionToolCall

            An array of tool calls the run step was involved in. These can be associated with one of three types of tools: code_interpreter, file_search, or function.

            • CodeInterpreterToolCall object { id, code_interpreter, type }

              Details of the Code Interpreter tool call the run step was involved in.

              • id: string

                The ID of the tool call.

              • code_interpreter: object { input, outputs }

                The Code Interpreter tool call definition.

                • input: string

                  The input to the Code Interpreter tool call.

                • outputs: array of object { logs, type } or object { image, type }

                  The outputs from the Code Interpreter tool call. Code Interpreter can output one or more items, including text (logs) or images (image). Each of these are represented by a different object type.

                  • CodeInterpreterLogOutput object { logs, type }

                    Text output from the Code Interpreter tool call as part of a run step.

                    • logs: string

                      The text output from the Code Interpreter tool call.

                    • type: "logs"

                      Always logs.

                      • "logs"

                  • CodeInterpreterImageOutput object { image, type }

                    • image: object { file_id }

                      • file_id: string

                        The file ID of the image.

                    • type: "image"

                      Always image.

                      • "image"

              • type: "code_interpreter"

                The type of tool call. This is always going to be code_interpreter for this type of tool call.

                • "code_interpreter"

            • FileSearchToolCall object { id, file_search, type }

              • id: string

                The ID of the tool call object.

              • file_search: object { ranking_options, results }

                For now, this is always going to be an empty object.

                • ranking_options: optional object { ranker, score_threshold }

                  The ranking options for the file search.

                  • ranker: "auto" or "default_2024_08_21"

                    The ranker to use for the file search. If not specified will use the auto ranker.

                    • "auto"

                    • "default_2024_08_21"

                  • score_threshold: number

                    The score threshold for the file search. All values must be a floating point number between 0 and 1.

                • results: optional array of object { file_id, file_name, score, content }

                  The results of the file search.

                  • file_id: string

                    The ID of the file that result was found in.

                  • file_name: string

                    The name of the file that result was found in.

                  • score: number

                    The score of the result. All values must be a floating point number between 0 and 1.

                  • content: optional array of object { text, type }

                    The content of the result that was found. The content is only included if requested via the include query parameter.

                    • text: optional string

                      The text content of the file.

                    • type: optional "text"

                      The type of the content.

                      • "text"

              • type: "file_search"

                The type of tool call. This is always going to be file_search for this type of tool call.

                • "file_search"

            • FunctionToolCall object { id, function, type }

              • id: string

                The ID of the tool call object.

              • function: object { arguments, name, output }

                The definition of the function that was called.

                • arguments: string

                  The arguments passed to the function.

                • name: string

                  The name of the function.

                • output: string or null

                  The output of the function. This will be null if the outputs have not been submitted yet.

              • type: "function"

                The type of tool call. This is always going to be function for this type of tool call.

                • "function"

          • type: "tool_calls"

            Always tool_calls.

            • "tool_calls"

      • thread_id: string

        The ID of the thread that was run.

      • type: "message_creation" or "tool_calls"

        The type of run step, which can be either message_creation or tool_calls.

        • "message_creation"

        • "tool_calls"

      • usage: object { completion_tokens, prompt_tokens, total_tokens } or null

        Usage statistics related to the run step. This value will be null while the run step's status is in_progress.

        • completion_tokens: number

          Number of completion tokens used over the course of the run step.

        • prompt_tokens: number

          Number of prompt tokens used over the course of the run step.

        • total_tokens: number

          Total number of tokens used (prompt + completion).

    • event: "thread.run.step.created"

      • "thread.run.step.created"

  • ThreadRunStepInProgress object { data, event }

    Occurs when a run step moves to an in_progress state.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.in_progress"

      • "thread.run.step.in_progress"

  • ThreadRunStepDelta object { data, event }

    Occurs when parts of a run step are being streamed.

    • data: RunStepDeltaEvent

      Represents a run step delta i.e. any changed fields on a run step during streaming.

      • id: string

        The identifier of the run step, which can be referenced in API endpoints.

      • delta: object { step_details }

        The delta containing the fields that have changed on the run step.

        • step_details: optional RunStepDeltaMessageDelta or ToolCallDeltaObject

          The details of the run step.

          • RunStepDeltaMessageDelta object { type, message_creation }

            Details of the message creation by the run step.

            • type: "message_creation"

              Always message_creation.

              • "message_creation"

            • message_creation: optional object { message_id }

              • message_id: optional string

                The ID of the message that was created by this run step.

          • ToolCallDeltaObject object { type, tool_calls }

            Details of the tool call.

            • type: "tool_calls"

              Always tool_calls.

              • "tool_calls"

            • tool_calls: optional array of CodeInterpreterToolCallDelta or FileSearchToolCallDelta or FunctionToolCallDelta

              An array of tool calls the run step was involved in. These can be associated with one of three types of tools: code_interpreter, file_search, or function.

              • CodeInterpreterToolCallDelta object { index, type, id, code_interpreter }

                Details of the Code Interpreter tool call the run step was involved in.

                • index: number

                  The index of the tool call in the tool calls array.

                • type: "code_interpreter"

                  The type of tool call. This is always going to be code_interpreter for this type of tool call.

                  • "code_interpreter"

                • id: optional string

                  The ID of the tool call.

                • code_interpreter: optional object { input, outputs }

                  The Code Interpreter tool call definition.

                  • input: optional string

                    The input to the Code Interpreter tool call.

                  • outputs: optional array of CodeInterpreterLogs or CodeInterpreterOutputImage

                    The outputs from the Code Interpreter tool call. Code Interpreter can output one or more items, including text (logs) or images (image). Each of these are represented by a different object type.

                    • CodeInterpreterLogs object { index, type, logs }

                      Text output from the Code Interpreter tool call as part of a run step.

                      • index: number

                        The index of the output in the outputs array.

                      • type: "logs"

                        Always logs.

                        • "logs"

                      • logs: optional string

                        The text output from the Code Interpreter tool call.

                    • CodeInterpreterOutputImage object { index, type, image }

                      • index: number

                        The index of the output in the outputs array.

                      • type: "image"

                        Always image.

                        • "image"

                      • image: optional object { file_id }

                        • file_id: optional string

                          The file ID of the image.

              • FileSearchToolCallDelta object { file_search, index, type, id }

                • file_search: unknown

                  For now, this is always going to be an empty object.

                • index: number

                  The index of the tool call in the tool calls array.

                • type: "file_search"

                  The type of tool call. This is always going to be file_search for this type of tool call.

                  • "file_search"

                • id: optional string

                  The ID of the tool call object.

              • FunctionToolCallDelta object { index, type, id, function }

                • index: number

                  The index of the tool call in the tool calls array.

                • type: "function"

                  The type of tool call. This is always going to be function for this type of tool call.

                  • "function"

                • id: optional string

                  The ID of the tool call object.

                • function: optional object { arguments, name, output }

                  The definition of the function that was called.

                  • arguments: optional string

                    The arguments passed to the function.

                  • name: optional string

                    The name of the function.

                  • output: optional string or null

                    The output of the function. This will be null if the outputs have not been submitted yet.

      • object: "thread.run.step.delta"

        The object type, which is always thread.run.step.delta.

        • "thread.run.step.delta"

    • event: "thread.run.step.delta"

      • "thread.run.step.delta"

  • ThreadRunStepCompleted object { data, event }

    Occurs when a run step is completed.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.completed"

      • "thread.run.step.completed"

  • ThreadRunStepFailed object { data, event }

    Occurs when a run step fails.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.failed"

      • "thread.run.step.failed"

  • ThreadRunStepCancelled object { data, event }

    Occurs when a run step is cancelled.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.cancelled"

      • "thread.run.step.cancelled"

  • ThreadRunStepExpired object { data, event }

    Occurs when a run step expires.

    • data: RunStep

      Represents a step in execution of a run.

    • event: "thread.run.step.expired"

      • "thread.run.step.expired"

Run Stream Event

• RunStreamEvent = object { data, event } or object { data, event } or object { data, event } or 7 more

  Occurs when a new run is created.

  • ThreadRunCreated object { data, event }

    Occurs when a new run is created.

    • data: Run

      Represents an execution run on a thread.

      • id: string

        The identifier, which can be referenced in API endpoints.

      • assistant_id: string

        The ID of the assistant used for execution of this run.

      • cancelled_at: number or null

        The Unix timestamp (in seconds) for when the run was cancelled.

      • completed_at: number or null

        The Unix timestamp (in seconds) for when the run was completed.

      • created_at: number

        The Unix timestamp (in seconds) for when the run was created.

      • expires_at: number or null

        The Unix timestamp (in seconds) for when the run will expire.

      • failed_at: number or null

        The Unix timestamp (in seconds) for when the run failed.

      • incomplete_details: object { reason } or null

        Details on why the run is incomplete. Will be null if the run is not incomplete.

        • reason: optional "max_completion_tokens" or "max_prompt_tokens"

          The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.

          • "max_completion_tokens"

          • "max_prompt_tokens"

      • instructions: string

        The instructions that the assistant used for this run.

      • last_error: object { code, message } or null

        The last error associated with this run. Will be null if there are no errors.

        • code: "server_error" or "rate_limit_exceeded" or "invalid_prompt"

          One of server_error, rate_limit_exceeded, or invalid_prompt.

          • "server_error"

          • "rate_limit_exceeded"

          • "invalid_prompt"

        • message: string

          A human-readable description of the error.

      • max_completion_tokens: number or null

        The maximum number of completion tokens specified to have been used over the course of the run.

      • max_prompt_tokens: number or null

        The maximum number of prompt tokens specified to have been used over the course of the run.

      • 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 format, and querying for objects via API or the dashboard.

        Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

      • model: string

        The model that the assistant used for this run.

      • object: "thread.run"

        The object type, which is always thread.run.

        • "thread.run"

      • parallel_tool_calls: boolean

        Whether to enable parallel function calling during tool use.

      • required_action: object { submit_tool_outputs, type } or null

        Details on the action required to continue the run. Will be null if no action is required.

        • submit_tool_outputs: object { tool_calls }

          Details on the tool outputs needed for this run to continue.

          • tool_calls: array of RequiredActionFunctionToolCall

            A list of the relevant tool calls.

            • id: string

              The ID of the tool call. This ID must be referenced when you submit the tool outputs in using the Submit tool outputs to run endpoint.

            • function: object { arguments, name }

              The function definition.

              • arguments: string

                The arguments that the model expects you to pass to the function.

              • name: string

                The name of the function.

            • type: "function"

              The type of tool call the output is required for. For now, this is always function.

              • "function"

        • type: "submit_tool_outputs"

          For now, this is always submit_tool_outputs.

          • "submit_tool_outputs"

      • response_format: AssistantResponseFormatOption or null

        Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

        Setting to { "type": "json_schema", "json_schema": {...} } enables Structured Outputs which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide.

        Setting to { "type": "json_object" } enables JSON mode, which ensures the message the model generates is valid JSON.

        Important: when using JSON mode, you must also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if finish_reason="length", which indicates the generation exceeded max_tokens or the conversation exceeded the max context length.

        • "auto"

          auto is the default value

          • "auto"

        • ResponseFormatText object { type }

          Default response format. Used to generate text responses.

          • type: "text"

            The type of response format being defined. Always text.

            • "text"

        • ResponseFormatJSONObject object { type }

          JSON object response format. An older method of generating JSON responses. Using json_schema is recommended for models that support it. Note that the model will not generate JSON without a system or user message instructing it to do so.

          • type: "json_object"

            The type of response format being defined. Always json_object.

            • "json_object"

        • ResponseFormatJSONSchema object { json_schema, type }

          JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.

          • json_schema: object { name, description, schema, strict }

            Structured Outputs configuration options, including a JSON Schema.

            • name: string

              The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

            • description: optional string

              A description of what the response format is for, used by the model to determine how to respond in the format.

            • schema: optional map[unknown]

              The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.

            • 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 in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.

          • type: "json_schema"

            The type of response format being defined. Always json_schema.

            • "json_schema"

      • started_at: number or null

        The Unix timestamp (in seconds) for when the run was started.

      • status: "queued" or "in_progress" or "requires_action" or 6 more

        The status of the run, which can be either queued, in_progress, requires_action, cancelling, cancelled, failed, completed, incomplete, or expired.

        • "queued"

        • "in_progress"

        • "requires_action"

        • "cancelling"

        • "cancelled"

        • "failed"

        • "completed"

        • "incomplete"

        • "expired"

      • thread_id: string

        The ID of the thread that was executed on as a part of this run.

      • tool_choice: AssistantToolChoiceOption or null

        Controls which (if any) tool is called by the model. none means the model will not call any tools and instead generates a message. auto is the default value and means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools before responding to the user. Specifying a particular tool like {"type": "file_search"} or {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool.

        • "none" or "auto" or "required"

          none means the model will not call any tools and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools before responding to the user.

          • "none"

          • "auto"

          • "required"

        • AssistantToolChoice object { type, function }

          Specifies a tool the model should use. Use to force the model to call a specific tool.

          • type: "function" or "code_interpreter" or "file_search"

            The type of the tool. If type is function, the function name must be set

            • "function"

            • "code_interpreter"

            • "file_search"

          • function: optional AssistantToolChoiceFunction

            • name: string

              The name of the function to call.

      • tools: array of CodeInterpreterTool or FileSearchTool or FunctionTool

        The list of tools that the assistant used for this run.

        • CodeInterpreterTool object { type }

          • type: "code_interpreter"

            The type of tool being defined: code_interpreter

            • "code_interpreter"

        • FileSearchTool object { type, file_search }

          • type: "file_search"

            The type of tool being defined: file_search

            • "file_search"

          • file_search: optional object { max_num_results, ranking_options }

            Overrides for the file search tool.

            • max_num_results: optional number

              The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.

              Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.

            • ranking_options: optional object { score_threshold, ranker }

              The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.

              See the file search tool documentation for more information.

              • score_threshold: number

                The score threshold for the file search. All values must be a floating point number between 0 and 1.

              • ranker: optional "auto" or "default_2024_08_21"

                The ranker to use for the file search. If not specified will use the auto ranker.

                • "auto"

                • "default_2024_08_21"

        • FunctionTool object { function, type }

          • function: FunctionDefinition

            • name: string

              The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

            • description: optional string

              A description of what the function does, used by the model to choose when and how to call the function.

            • parameters: optional FunctionParameters

              The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.

              Omitting parameters defines a function with an empty parameter list.

            • strict: optional boolean or null

              Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.

          • type: "function"

            The type of tool being defined: function

            • "function"

      • truncation_strategy: object { type, last_messages } or null

        Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

        • type: "auto" or "last_messages"

          The truncation strategy to use for the thread. The default is auto. If set to last_messages, the thread will be truncated to the n most recent messages in the thread. When set to auto, messages in the middle of the thread will be dropped to fit the context length of the model, max_prompt_tokens.

          • "auto"

          • "last_messages"

        • last_messages: optional number or null

          The number of most recent messages from the thread when constructing the context for the run.

      • usage: object { completion_tokens, prompt_tokens, total_tokens } or null

        Usage statistics related to the run. This value will be null if the run is not in a terminal state (i.e. in_progress, queued, etc.).

        • completion_tokens: number

          Number of completion tokens used over the course of the run.

        • prompt_tokens: number

          Number of prompt tokens used over the course of the run.

        • total_tokens: number

          Total number of tokens used (prompt + completion).

      • temperature: optional number or null

        The sampling temperature used for this run. If not set, defaults to 1.

      • top_p: optional number or null

        The nucleus sampling value used for this run. If not set, defaults to 1.

    • event: "thread.run.created"

      • "thread.run.created"

  • ThreadRunQueued object { data, event }

    Occurs when a run moves to a queued status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.queued"

      • "thread.run.queued"

  • ThreadRunInProgress object { data, event }

    Occurs when a run moves to an in_progress status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.in_progress"

      • "thread.run.in_progress"

  • ThreadRunRequiresAction object { data, event }

    Occurs when a run moves to a requires_action status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.requires_action"

      • "thread.run.requires_action"

  • ThreadRunCompleted object { data, event }

    Occurs when a run is completed.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.completed"

      • "thread.run.completed"

  • ThreadRunIncomplete object { data, event }

    Occurs when a run ends with status incomplete.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.incomplete"

      • "thread.run.incomplete"

  • ThreadRunFailed object { data, event }

    Occurs when a run fails.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.failed"

      • "thread.run.failed"

  • ThreadRunCancelling object { data, event }

    Occurs when a run moves to a cancelling status.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.cancelling"

      • "thread.run.cancelling"

  • ThreadRunCancelled object { data, event }

    Occurs when a run is cancelled.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.cancelled"

      • "thread.run.cancelled"

  • ThreadRunExpired object { data, event }

    Occurs when a run expires.

    • data: Run

      Represents an execution run on a thread.

    • event: "thread.run.expired"

      • "thread.run.expired"

Thread Stream Event

• ThreadStreamEvent object { data, event, enabled }

  Occurs when a new thread is created.

  • data: Thread

    Represents a thread that contains messages.

    • id: string

      The identifier, which can be referenced in API endpoints.

    • created_at: number

      The Unix timestamp (in seconds) for when the thread was created.

    • 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 format, and querying for objects via API or the dashboard.

      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.

    • object: "thread"

      The object type, which is always thread.

      • "thread"

    • tool_resources: object { code_interpreter, file_search } or null

      A set of resources that are made available to the assistant's tools in this thread. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

      • code_interpreter: optional object { file_ids }

        • file_ids: optional array of string

          A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.

      • file_search: optional object { vector_store_ids }

        • vector_store_ids: optional array of string

          The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.

  • event: "thread.created"

    • "thread.created"

  • enabled: optional boolean

    Whether to enable input audio transcription.