Assistants

List assistants

AssistantListPage beta().assistants().list(AssistantListParamsparams = AssistantListParams.none(), RequestOptionsrequestOptions = RequestOptions.none())

get /assistants

List assistants

Parameters

AssistantListParams params
- Optional<String> after
  
  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.
- Optional<String> before
  
  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.
- Optional<Long> limit
  
  A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.
- Optional<Order> order
  
  Sort order by the created_at timestamp of the objects. asc for ascending order and desc for descending order.
  - ASC("asc")
  - DESC("desc")

Returns

class Assistant:

Represents an assistant that can call the model and use tools.
- String id
  
  The identifier, which can be referenced in API endpoints.
- long createdAt
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Optional<String> description
  
  The description of the assistant. The maximum length is 512 characters.
- Optional<String> instructions
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Optional<Metadata> metadata
  
  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.
- String model
  
  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.
- Optional<String> name
  
  The name of the assistant. The maximum length is 256 characters.
- JsonValue; object_ "assistant"constant
  
  The object type, which is always assistant.
  - ASSISTANT("assistant")
- List<JsonValue> tools
  
  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.
- Optional<AssistantResponseFormatOption> responseFormat
  
  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.
  - JsonValue;
    - AUTO("auto")
  - class ResponseFormatText:
    
    Default response format. Used to generate text responses.
    - JsonValue; type "text"constant
      
      The type of response format being defined. Always text.
      - TEXT("text")
  - class ResponseFormatJsonObject:
    
    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.
    - JsonValue; type "json_object"constant
      
      The type of response format being defined. Always json_object.
      - JSON_OBJECT("json_object")
  - class ResponseFormatJsonSchema:
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JsonSchema jsonSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - String name
        
        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.
      - Optional<String> description
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Optional<Schema> schema
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Optional<Boolean> strict
        
        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.
    - JsonValue; type "json_schema"constant
      
      The type of response format being defined. Always json_schema.
      - JSON_SCHEMA("json_schema")
- Optional<Double> temperature
  
  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.
- Optional<ToolResources> toolResources
  
  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.
  - Optional<CodeInterpreter> codeInterpreter
    - Optional<List<String>> fileIds
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - Optional<FileSearch> fileSearch
    - Optional<List<String>> vectorStoreIds
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- Optional<Double> topP
  
  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

package com.openai.example;

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.beta.assistants.AssistantListPage;
import com.openai.models.beta.assistants.AssistantListParams;

public final class Main {
    private Main() {}

    public static void main(String[] args) {
        OpenAIClient client = OpenAIOkHttpClient.fromEnv();

        AssistantListPage page = client.beta().assistants().list();
    }
}

Response

{
  "data": [
    {
      "id": "id",
      "created_at": 0,
      "description": "description",
      "instructions": "instructions",
      "metadata": {
        "foo": "string"
      },
      "model": "model",
      "name": "name",
      "object": "assistant",
      "tools": [
        {}
      ],
      "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"
}

Create assistant

Assistant beta().assistants().create(AssistantCreateParamsparams, RequestOptionsrequestOptions = RequestOptions.none())

post /assistants

Create assistant

Parameters

AssistantCreateParams params
- ChatModel model
  
  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.
- Optional<String> description
  
  The description of the assistant. The maximum length is 512 characters.
- Optional<String> instructions
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Optional<Metadata> metadata
  
  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.
- Optional<String> name
  
  The name of the assistant. The maximum length is 256 characters.
- Optional<ReasoningEffort> reasoningEffort
  
  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.
- Optional<AssistantResponseFormatOption> responseFormat
  
  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.
- Optional<Double> temperature
  
  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.
- Optional<ToolResources> toolResources
  
  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.
  - Optional<CodeInterpreter> codeInterpreter
    - Optional<List<String>> fileIds
      
      A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
  - Optional<FileSearch> fileSearch
    - Optional<List<String>> vectorStoreIds
      
      The vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
    - Optional<List<VectorStore>> vectorStores
      
      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.
      - Optional<ChunkingStrategy> chunkingStrategy
        
        The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.
        
        JsonValue;
        
        JsonValue; type "auto"constant
        
        Always auto.
        
        AUTO("auto")
        
        class Static:
        
        InnerStatic static_
        
        long chunkOverlapTokens
        
        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.
        
        long maxChunkSizeTokens
        
        The maximum number of tokens in each chunk. The default value is 800. The minimum value is 100 and the maximum value is 4096.
        
        JsonValue; type "static"constant
        
        Always static.
        
        STATIC("static")
      - Optional<List<String>> fileIds
        
        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.
      - Optional<Metadata> metadata
        
        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.
- Optional<List<JsonValue>> tools
  
  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.
- Optional<Double> topP
  
  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

class Assistant:

Represents an assistant that can call the model and use tools.
- String id
  
  The identifier, which can be referenced in API endpoints.
- long createdAt
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Optional<String> description
  
  The description of the assistant. The maximum length is 512 characters.
- Optional<String> instructions
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Optional<Metadata> metadata
  
  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.
- String model
  
  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.
- Optional<String> name
  
  The name of the assistant. The maximum length is 256 characters.
- JsonValue; object_ "assistant"constant
  
  The object type, which is always assistant.
  - ASSISTANT("assistant")
- List<JsonValue> tools
  
  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.
- Optional<AssistantResponseFormatOption> responseFormat
  
  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.
  - JsonValue;
    - AUTO("auto")
  - class ResponseFormatText:
    
    Default response format. Used to generate text responses.
    - JsonValue; type "text"constant
      
      The type of response format being defined. Always text.
      - TEXT("text")
  - class ResponseFormatJsonObject:
    
    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.
    - JsonValue; type "json_object"constant
      
      The type of response format being defined. Always json_object.
      - JSON_OBJECT("json_object")
  - class ResponseFormatJsonSchema:
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JsonSchema jsonSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - String name
        
        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.
      - Optional<String> description
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Optional<Schema> schema
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Optional<Boolean> strict
        
        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.
    - JsonValue; type "json_schema"constant
      
      The type of response format being defined. Always json_schema.
      - JSON_SCHEMA("json_schema")
- Optional<Double> temperature
  
  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.
- Optional<ToolResources> toolResources
  
  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.
  - Optional<CodeInterpreter> codeInterpreter
    - Optional<List<String>> fileIds
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - Optional<FileSearch> fileSearch
    - Optional<List<String>> vectorStoreIds
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- Optional<Double> topP
  
  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

package com.openai.example;

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.ChatModel;
import com.openai.models.beta.assistants.Assistant;
import com.openai.models.beta.assistants.AssistantCreateParams;

public final class Main {
    private Main() {}

    public static void main(String[] args) {
        OpenAIClient client = OpenAIOkHttpClient.fromEnv();

        AssistantCreateParams params = AssistantCreateParams.builder()
            .model(ChatModel.GPT_4O)
            .build();
        Assistant assistant = client.beta().assistants().create(params);
    }
}

Response

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

Retrieve assistant

Assistant beta().assistants().retrieve(AssistantRetrieveParamsparams = AssistantRetrieveParams.none(), RequestOptionsrequestOptions = RequestOptions.none())

get /assistants/{assistant_id}

Retrieve assistant

Parameters

AssistantRetrieveParams params
- Optional<String> assistantId

Returns

class Assistant:

Represents an assistant that can call the model and use tools.
- String id
  
  The identifier, which can be referenced in API endpoints.
- long createdAt
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Optional<String> description
  
  The description of the assistant. The maximum length is 512 characters.
- Optional<String> instructions
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Optional<Metadata> metadata
  
  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.
- String model
  
  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.
- Optional<String> name
  
  The name of the assistant. The maximum length is 256 characters.
- JsonValue; object_ "assistant"constant
  
  The object type, which is always assistant.
  - ASSISTANT("assistant")
- List<JsonValue> tools
  
  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.
- Optional<AssistantResponseFormatOption> responseFormat
  
  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.
  - JsonValue;
    - AUTO("auto")
  - class ResponseFormatText:
    
    Default response format. Used to generate text responses.
    - JsonValue; type "text"constant
      
      The type of response format being defined. Always text.
      - TEXT("text")
  - class ResponseFormatJsonObject:
    
    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.
    - JsonValue; type "json_object"constant
      
      The type of response format being defined. Always json_object.
      - JSON_OBJECT("json_object")
  - class ResponseFormatJsonSchema:
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JsonSchema jsonSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - String name
        
        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.
      - Optional<String> description
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Optional<Schema> schema
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Optional<Boolean> strict
        
        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.
    - JsonValue; type "json_schema"constant
      
      The type of response format being defined. Always json_schema.
      - JSON_SCHEMA("json_schema")
- Optional<Double> temperature
  
  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.
- Optional<ToolResources> toolResources
  
  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.
  - Optional<CodeInterpreter> codeInterpreter
    - Optional<List<String>> fileIds
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - Optional<FileSearch> fileSearch
    - Optional<List<String>> vectorStoreIds
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- Optional<Double> topP
  
  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

package com.openai.example;

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.beta.assistants.Assistant;
import com.openai.models.beta.assistants.AssistantRetrieveParams;

public final class Main {
    private Main() {}

    public static void main(String[] args) {
        OpenAIClient client = OpenAIOkHttpClient.fromEnv();

        Assistant assistant = client.beta().assistants().retrieve("assistant_id");
    }
}

Response

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

Modify assistant

Assistant beta().assistants().update(AssistantUpdateParamsparams = AssistantUpdateParams.none(), RequestOptionsrequestOptions = RequestOptions.none())

post /assistants/{assistant_id}

Modify assistant

Parameters

AssistantUpdateParams params
- Optional<String> assistantId
- Optional<String> description
  
  The description of the assistant. The maximum length is 512 characters.
- Optional<String> instructions
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Optional<Metadata> metadata
  
  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.
- Optional<Model> model
  
  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")
  - GPT_5_MINI("gpt-5-mini")
  - GPT_5_NANO("gpt-5-nano")
  - GPT_5_2025_08_07("gpt-5-2025-08-07")
  - GPT_5_MINI_2025_08_07("gpt-5-mini-2025-08-07")
  - GPT_5_NANO_2025_08_07("gpt-5-nano-2025-08-07")
  - GPT_4_1("gpt-4.1")
  - GPT_4_1_MINI("gpt-4.1-mini")
  - GPT_4_1_NANO("gpt-4.1-nano")
  - GPT_4_1_2025_04_14("gpt-4.1-2025-04-14")
  - GPT_4_1_MINI_2025_04_14("gpt-4.1-mini-2025-04-14")
  - GPT_4_1_NANO_2025_04_14("gpt-4.1-nano-2025-04-14")
  - O3_MINI("o3-mini")
  - O3_MINI_2025_01_31("o3-mini-2025-01-31")
  - O1("o1")
  - O1_2024_12_17("o1-2024-12-17")
  - GPT_4O("gpt-4o")
  - GPT_4O_2024_11_20("gpt-4o-2024-11-20")
  - GPT_4O_2024_08_06("gpt-4o-2024-08-06")
  - GPT_4O_2024_05_13("gpt-4o-2024-05-13")
  - GPT_4O_MINI("gpt-4o-mini")
  - GPT_4O_MINI_2024_07_18("gpt-4o-mini-2024-07-18")
  - GPT_4_5_PREVIEW("gpt-4.5-preview")
  - GPT_4_5_PREVIEW_2025_02_27("gpt-4.5-preview-2025-02-27")
  - GPT_4_TURBO("gpt-4-turbo")
  - GPT_4_TURBO_2024_04_09("gpt-4-turbo-2024-04-09")
  - GPT_4_0125_PREVIEW("gpt-4-0125-preview")
  - GPT_4_TURBO_PREVIEW("gpt-4-turbo-preview")
  - GPT_4_1106_PREVIEW("gpt-4-1106-preview")
  - GPT_4_VISION_PREVIEW("gpt-4-vision-preview")
  - GPT_4("gpt-4")
  - GPT_4_0314("gpt-4-0314")
  - GPT_4_0613("gpt-4-0613")
  - GPT_4_32K("gpt-4-32k")
  - GPT_4_32K_0314("gpt-4-32k-0314")
  - GPT_4_32K_0613("gpt-4-32k-0613")
  - GPT_3_5_TURBO("gpt-3.5-turbo")
  - GPT_3_5_TURBO_16K("gpt-3.5-turbo-16k")
  - GPT_3_5_TURBO_0613("gpt-3.5-turbo-0613")
  - GPT_3_5_TURBO_1106("gpt-3.5-turbo-1106")
  - GPT_3_5_TURBO_0125("gpt-3.5-turbo-0125")
  - GPT_3_5_TURBO_16K_0613("gpt-3.5-turbo-16k-0613")
- Optional<String> name
  
  The name of the assistant. The maximum length is 256 characters.
- Optional<ReasoningEffort> reasoningEffort
  
  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.
- Optional<AssistantResponseFormatOption> responseFormat
  
  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.
- Optional<Double> temperature
  
  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.
- Optional<ToolResources> toolResources
  
  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.
  - Optional<CodeInterpreter> codeInterpreter
    - Optional<List<String>> fileIds
      
      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.
  - Optional<FileSearch> fileSearch
    - Optional<List<String>> vectorStoreIds
      
      Overrides the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- Optional<List<JsonValue>> tools
  
  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.
- Optional<Double> topP
  
  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

class Assistant:

Represents an assistant that can call the model and use tools.
- String id
  
  The identifier, which can be referenced in API endpoints.
- long createdAt
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Optional<String> description
  
  The description of the assistant. The maximum length is 512 characters.
- Optional<String> instructions
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Optional<Metadata> metadata
  
  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.
- String model
  
  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.
- Optional<String> name
  
  The name of the assistant. The maximum length is 256 characters.
- JsonValue; object_ "assistant"constant
  
  The object type, which is always assistant.
  - ASSISTANT("assistant")
- List<JsonValue> tools
  
  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.
- Optional<AssistantResponseFormatOption> responseFormat
  
  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.
  - JsonValue;
    - AUTO("auto")
  - class ResponseFormatText:
    
    Default response format. Used to generate text responses.
    - JsonValue; type "text"constant
      
      The type of response format being defined. Always text.
      - TEXT("text")
  - class ResponseFormatJsonObject:
    
    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.
    - JsonValue; type "json_object"constant
      
      The type of response format being defined. Always json_object.
      - JSON_OBJECT("json_object")
  - class ResponseFormatJsonSchema:
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JsonSchema jsonSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - String name
        
        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.
      - Optional<String> description
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Optional<Schema> schema
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Optional<Boolean> strict
        
        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.
    - JsonValue; type "json_schema"constant
      
      The type of response format being defined. Always json_schema.
      - JSON_SCHEMA("json_schema")
- Optional<Double> temperature
  
  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.
- Optional<ToolResources> toolResources
  
  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.
  - Optional<CodeInterpreter> codeInterpreter
    - Optional<List<String>> fileIds
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - Optional<FileSearch> fileSearch
    - Optional<List<String>> vectorStoreIds
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- Optional<Double> topP
  
  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

package com.openai.example;

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.beta.assistants.Assistant;
import com.openai.models.beta.assistants.AssistantUpdateParams;

public final class Main {
    private Main() {}

    public static void main(String[] args) {
        OpenAIClient client = OpenAIOkHttpClient.fromEnv();

        Assistant assistant = client.beta().assistants().update("assistant_id");
    }
}

Response

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

Delete assistant

AssistantDeleted beta().assistants().delete(AssistantDeleteParamsparams = AssistantDeleteParams.none(), RequestOptionsrequestOptions = RequestOptions.none())

delete /assistants/{assistant_id}

Delete assistant

Parameters

AssistantDeleteParams params
- Optional<String> assistantId

Returns

class AssistantDeleted:
- String id
- boolean deleted
- JsonValue; object_ "assistant.deleted"constant
  - ASSISTANT_DELETED("assistant.deleted")

Example

package com.openai.example;

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.beta.assistants.AssistantDeleteParams;
import com.openai.models.beta.assistants.AssistantDeleted;

public final class Main {
    private Main() {}

    public static void main(String[] args) {
        OpenAIClient client = OpenAIOkHttpClient.fromEnv();

        AssistantDeleted assistantDeleted = client.beta().assistants().delete("assistant_id");
    }
}

Response

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

Domain Types

Assistant

class Assistant:

Represents an assistant that can call the model and use tools.
- String id
  
  The identifier, which can be referenced in API endpoints.
- long createdAt
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Optional<String> description
  
  The description of the assistant. The maximum length is 512 characters.
- Optional<String> instructions
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Optional<Metadata> metadata
  
  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.
- String model
  
  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.
- Optional<String> name
  
  The name of the assistant. The maximum length is 256 characters.
- JsonValue; object_ "assistant"constant
  
  The object type, which is always assistant.
  - ASSISTANT("assistant")
- List<JsonValue> tools
  
  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.
- Optional<AssistantResponseFormatOption> responseFormat
  
  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.
  - JsonValue;
    - AUTO("auto")
  - class ResponseFormatText:
    
    Default response format. Used to generate text responses.
    - JsonValue; type "text"constant
      
      The type of response format being defined. Always text.
      - TEXT("text")
  - class ResponseFormatJsonObject:
    
    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.
    - JsonValue; type "json_object"constant
      
      The type of response format being defined. Always json_object.
      - JSON_OBJECT("json_object")
  - class ResponseFormatJsonSchema:
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JsonSchema jsonSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - String name
        
        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.
      - Optional<String> description
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Optional<Schema> schema
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Optional<Boolean> strict
        
        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.
    - JsonValue; type "json_schema"constant
      
      The type of response format being defined. Always json_schema.
      - JSON_SCHEMA("json_schema")
- Optional<Double> temperature
  
  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.
- Optional<ToolResources> toolResources
  
  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.
  - Optional<CodeInterpreter> codeInterpreter
    - Optional<List<String>> fileIds
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - Optional<FileSearch> fileSearch
    - Optional<List<String>> vectorStoreIds
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- Optional<Double> topP
  
  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

class AssistantDeleted:
- String id
- boolean deleted
- JsonValue; object_ "assistant.deleted"constant
  - ASSISTANT_DELETED("assistant.deleted")

Assistant Stream Event

class AssistantStreamEvent: A class that can be one of several variants.union

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
  - Thread data
    
    Represents a thread that contains messages.
    - String id
      
      The identifier, which can be referenced in API endpoints.
    - long createdAt
      
      The Unix timestamp (in seconds) for when the thread was created.
    - Optional<Metadata> metadata
      
      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.
    - JsonValue; object_ "thread"constant
      
      The object type, which is always thread.
      - THREAD("thread")
    - Optional<ToolResources> toolResources
      
      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.
      - Optional<CodeInterpreter> codeInterpreter
        
        Optional<List<String>> fileIds
        
        A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
      - Optional<FileSearch> fileSearch
        
        Optional<List<String>> vectorStoreIds
        
        The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.
  - JsonValue; event "thread.created"constant
    - THREAD_CREATED("thread.created")
  - Optional<Boolean> enabled
    
    Whether to enable input audio transcription.
- ThreadRunCreated
  - Run data
    
    Represents an execution run on a thread.
    - String id
      
      The identifier, which can be referenced in API endpoints.
    - String assistantId
      
      The ID of the assistant used for execution of this run.
    - Optional<Long> cancelledAt
      
      The Unix timestamp (in seconds) for when the run was cancelled.
    - Optional<Long> completedAt
      
      The Unix timestamp (in seconds) for when the run was completed.
    - long createdAt
      
      The Unix timestamp (in seconds) for when the run was created.
    - Optional<Long> expiresAt
      
      The Unix timestamp (in seconds) for when the run will expire.
    - Optional<Long> failedAt
      
      The Unix timestamp (in seconds) for when the run failed.
    - Optional<IncompleteDetails> incompleteDetails
      
      Details on why the run is incomplete. Will be null if the run is not incomplete.
      - Optional<Reason> reason
        
        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_completion_tokens")
        
        MAX_PROMPT_TOKENS("max_prompt_tokens")
    - String instructions
      
      The instructions that the assistant used for this run.
    - Optional<LastError> lastError
      
      The last error associated with this run. Will be null if there are no errors.
      - Code code
        
        One of server_error, rate_limit_exceeded, or invalid_prompt.
        
        SERVER_ERROR("server_error")
        
        RATE_LIMIT_EXCEEDED("rate_limit_exceeded")
        
        INVALID_PROMPT("invalid_prompt")
      - String message
        
        A human-readable description of the error.
    - Optional<Long> maxCompletionTokens
      
      The maximum number of completion tokens specified to have been used over the course of the run.
    - Optional<Long> maxPromptTokens
      
      The maximum number of prompt tokens specified to have been used over the course of the run.
    - Optional<Metadata> metadata
      
      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.
    - String model
      
      The model that the assistant used for this run.
    - JsonValue; object_ "thread.run"constant
      
      The object type, which is always thread.run.
      - THREAD_RUN("thread.run")
    - boolean parallelToolCalls
      
      Whether to enable parallel function calling during tool use.
    - Optional<RequiredAction> requiredAction
      
      Details on the action required to continue the run. Will be null if no action is required.
      - SubmitToolOutputs submitToolOutputs
        
        Details on the tool outputs needed for this run to continue.
        
        List<RequiredActionFunctionToolCall> toolCalls
        
        A list of the relevant tool calls.
        
        String id
        
        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 function
        
        The function definition.
        
        String arguments
        
        The arguments that the model expects you to pass to the function.
        
        String name
        
        The name of the function.
        
        JsonValue; type "function"constant
        
        The type of tool call the output is required for. For now, this is always function.
        
        FUNCTION("function")
      - JsonValue; type "submit_tool_outputs"constant
        
        For now, this is always submit_tool_outputs.
        
        SUBMIT_TOOL_OUTPUTS("submit_tool_outputs")
    - Optional<AssistantResponseFormatOption> responseFormat
      
      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.
      - JsonValue;
        
        AUTO("auto")
      - class ResponseFormatText:
        
        Default response format. Used to generate text responses.
        
        JsonValue; type "text"constant
        
        The type of response format being defined. Always text.
        
        TEXT("text")
      - class ResponseFormatJsonObject:
        
        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.
        
        JsonValue; type "json_object"constant
        
        The type of response format being defined. Always json_object.
        
        JSON_OBJECT("json_object")
      - class ResponseFormatJsonSchema:
        
        JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
        
        JsonSchema jsonSchema
        
        Structured Outputs configuration options, including a JSON Schema.
        
        String name
        
        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.
        
        Optional<String> description
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
        
        Optional<Schema> schema
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
        
        Optional<Boolean> strict
        
        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.
        
        JsonValue; type "json_schema"constant
        
        The type of response format being defined. Always json_schema.
        
        JSON_SCHEMA("json_schema")
    - Optional<Long> startedAt
      
      The Unix timestamp (in seconds) for when the run was started.
    - JsonValue status
    - String threadId
      
      The ID of the thread that was executed on as a part of this run.
    - Optional<AssistantToolChoiceOption> toolChoice
      
      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.
      - Auto
        
        NONE("none")
        
        AUTO("auto")
        
        REQUIRED("required")
      - class AssistantToolChoice:
        
        Specifies a tool the model should use. Use to force the model to call a specific tool.
        
        Type type
        
        The type of the tool. If type is function, the function name must be set
        
        FUNCTION("function")
        
        CODE_INTERPRETER("code_interpreter")
        
        FILE_SEARCH("file_search")
        
        Optional<AssistantToolChoiceFunction> function
        
        String name
        
        The name of the function to call.
    - List<JsonValue> tools
      
      The list of tools that the assistant used for this run.
    - Optional<TruncationStrategy> truncationStrategy
      
      Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.
      - Type type
        
        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("auto")
        
        LAST_MESSAGES("last_messages")
      - Optional<Long> lastMessages
        
        The number of most recent messages from the thread when constructing the context for the run.
    - Optional<Usage> usage
      
      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.).
      - long completionTokens
        
        Number of completion tokens used over the course of the run.
      - long promptTokens
        
        Number of prompt tokens used over the course of the run.
      - long totalTokens
        
        Total number of tokens used (prompt + completion).
    - Optional<Double> temperature
      
      The sampling temperature used for this run. If not set, defaults to 1.
    - Optional<Double> topP
      
      The nucleus sampling value used for this run. If not set, defaults to 1.
  - JsonValue; event "thread.run.created"constant
    - THREAD_RUN_CREATED("thread.run.created")
- ThreadRunQueued
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.queued"constant
    - THREAD_RUN_QUEUED("thread.run.queued")
- ThreadRunInProgress
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.in_progress"constant
    - THREAD_RUN_IN_PROGRESS("thread.run.in_progress")
- ThreadRunRequiresAction
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.requires_action"constant
    - THREAD_RUN_REQUIRES_ACTION("thread.run.requires_action")
- ThreadRunCompleted
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.completed"constant
    - THREAD_RUN_COMPLETED("thread.run.completed")
- ThreadRunIncomplete
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.incomplete"constant
    - THREAD_RUN_INCOMPLETE("thread.run.incomplete")
- ThreadRunFailed
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.failed"constant
    - THREAD_RUN_FAILED("thread.run.failed")
- ThreadRunCancelling
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.cancelling"constant
    - THREAD_RUN_CANCELLING("thread.run.cancelling")
- ThreadRunCancelled
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.cancelled"constant
    - THREAD_RUN_CANCELLED("thread.run.cancelled")
- ThreadRunExpired
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.expired"constant
    - THREAD_RUN_EXPIRED("thread.run.expired")
- ThreadRunStepCreated
  - RunStep data
    
    Represents a step in execution of a run.
    - String id
      
      The identifier of the run step, which can be referenced in API endpoints.
    - String assistantId
      
      The ID of the assistant associated with the run step.
    - Optional<Long> cancelledAt
      
      The Unix timestamp (in seconds) for when the run step was cancelled.
    - Optional<Long> completedAt
      
      The Unix timestamp (in seconds) for when the run step completed.
    - long createdAt
      
      The Unix timestamp (in seconds) for when the run step was created.
    - Optional<Long> expiredAt
      
      The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.
    - Optional<Long> failedAt
      
      The Unix timestamp (in seconds) for when the run step failed.
    - Optional<LastError> lastError
      
      The last error associated with this run step. Will be null if there are no errors.
      - Code code
        
        One of server_error or rate_limit_exceeded.
        
        SERVER_ERROR("server_error")
        
        RATE_LIMIT_EXCEEDED("rate_limit_exceeded")
      - String message
        
        A human-readable description of the error.
    - Optional<Metadata> metadata
      
      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.
    - JsonValue; object_ "thread.run.step"constant
      
      The object type, which is always thread.run.step.
      - THREAD_RUN_STEP("thread.run.step")
    - String runId
      
      The ID of the run that this run step is a part of.
    - Status status
      
      The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.
      - IN_PROGRESS("in_progress")
      - CANCELLED("cancelled")
      - FAILED("failed")
      - COMPLETED("completed")
      - EXPIRED("expired")
    - StepDetails stepDetails
      
      The details of the run step.
      - class MessageCreationStepDetails:
        
        Details of the message creation by the run step.
        
        MessageCreation messageCreation
        
        String messageId
        
        The ID of the message that was created by this run step.
        
        JsonValue; type "message_creation"constant
        
        Always message_creation.
        
        MESSAGE_CREATION("message_creation")
      - class ToolCallsStepDetails:
        
        Details of the tool call.
        
        List<JsonValue> toolCalls
        
        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.
        
        JsonValue; type "tool_calls"constant
        
        Always tool_calls.
        
        TOOL_CALLS("tool_calls")
    - String threadId
      
      The ID of the thread that was run.
    - Type type
      
      The type of run step, which can be either message_creation or tool_calls.
      - MESSAGE_CREATION("message_creation")
      - TOOL_CALLS("tool_calls")
    - Optional<Usage> usage
      
      Usage statistics related to the run step. This value will be null while the run step's status is in_progress.
      - long completionTokens
        
        Number of completion tokens used over the course of the run step.
      - long promptTokens
        
        Number of prompt tokens used over the course of the run step.
      - long totalTokens
        
        Total number of tokens used (prompt + completion).
  - JsonValue; event "thread.run.step.created"constant
    - THREAD_RUN_STEP_CREATED("thread.run.step.created")
- ThreadRunStepInProgress
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.in_progress"constant
    - THREAD_RUN_STEP_IN_PROGRESS("thread.run.step.in_progress")
- ThreadRunStepDelta
  - RunStepDeltaEvent data
    
    Represents a run step delta i.e. any changed fields on a run step during streaming.
    - String id
      
      The identifier of the run step, which can be referenced in API endpoints.
    - JsonValue delta
    - JsonValue; object_ "thread.run.step.delta"constant
      
      The object type, which is always thread.run.step.delta.
      - THREAD_RUN_STEP_DELTA("thread.run.step.delta")
  - JsonValue; event "thread.run.step.delta"constant
    - THREAD_RUN_STEP_DELTA("thread.run.step.delta")
- ThreadRunStepCompleted
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.completed"constant
    - THREAD_RUN_STEP_COMPLETED("thread.run.step.completed")
- ThreadRunStepFailed
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.failed"constant
    - THREAD_RUN_STEP_FAILED("thread.run.step.failed")
- ThreadRunStepCancelled
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.cancelled"constant
    - THREAD_RUN_STEP_CANCELLED("thread.run.step.cancelled")
- ThreadRunStepExpired
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.expired"constant
    - THREAD_RUN_STEP_EXPIRED("thread.run.step.expired")
- ThreadMessageCreated
  - Message data
    
    Represents a message within a thread.
    - String id
      
      The identifier, which can be referenced in API endpoints.
    - Optional<String> assistantId
      
      If applicable, the ID of the assistant that authored this message.
    - Optional<List<Attachment>> attachments
      
      A list of files attached to the message, and the tools they were added to.
      - Optional<String> fileId
        
        The ID of the file to attach to the message.
      - Optional<List<Tool>> tools
        
        The tools to add this file to.
        
        class CodeInterpreterTool:
        
        JsonValue; type "code_interpreter"constant
        
        The type of tool being defined: code_interpreter
        
        CODE_INTERPRETER("code_interpreter")
        
        JsonValue;
        
        JsonValue; type "file_search"constant
        
        The type of tool being defined: file_search
        
        FILE_SEARCH("file_search")
    - Optional<Long> completedAt
      
      The Unix timestamp (in seconds) for when the message was completed.
    - List<JsonValue> content
      
      The content of the message in array of text and/or images.
    - long createdAt
      
      The Unix timestamp (in seconds) for when the message was created.
    - Optional<Long> incompleteAt
      
      The Unix timestamp (in seconds) for when the message was marked as incomplete.
    - Optional<IncompleteDetails> incompleteDetails
      
      On an incomplete message, details about why the message is incomplete.
      - Reason reason
        
        The reason the message is incomplete.
        
        CONTENT_FILTER("content_filter")
        
        MAX_TOKENS("max_tokens")
        
        RUN_CANCELLED("run_cancelled")
        
        RUN_EXPIRED("run_expired")
        
        RUN_FAILED("run_failed")
    - Optional<Metadata> metadata
      
      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.
    - JsonValue; object_ "thread.message"constant
      
      The object type, which is always thread.message.
      - THREAD_MESSAGE("thread.message")
    - Role role
      
      The entity that produced the message. One of user or assistant.
      - USER("user")
      - ASSISTANT("assistant")
    - Optional<String> runId
      
      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 status
      
      The status of the message, which can be either in_progress, incomplete, or completed.
      - IN_PROGRESS("in_progress")
      - INCOMPLETE("incomplete")
      - COMPLETED("completed")
    - String threadId
      
      The thread ID that this message belongs to.
  - JsonValue; event "thread.message.created"constant
    - THREAD_MESSAGE_CREATED("thread.message.created")
- ThreadMessageInProgress
  - Message data
    
    Represents a message within a thread.
  - JsonValue; event "thread.message.in_progress"constant
    - THREAD_MESSAGE_IN_PROGRESS("thread.message.in_progress")
- ThreadMessageDelta
  - MessageDeltaEvent data
    
    Represents a message delta i.e. any changed fields on a message during streaming.
    - String id
      
      The identifier of the message, which can be referenced in API endpoints.
    - MessageDelta delta
      
      The delta containing the fields that have changed on the Message.
      - Optional<List<JsonValue>> content
        
        The content of the message in array of text and/or images.
      - Optional<Role> role
        
        The entity that produced the message. One of user or assistant.
        
        USER("user")
        
        ASSISTANT("assistant")
    - JsonValue; object_ "thread.message.delta"constant
      
      The object type, which is always thread.message.delta.
      - THREAD_MESSAGE_DELTA("thread.message.delta")
  - JsonValue; event "thread.message.delta"constant
    - THREAD_MESSAGE_DELTA("thread.message.delta")
- ThreadMessageCompleted
  - Message data
    
    Represents a message within a thread.
  - JsonValue; event "thread.message.completed"constant
    - THREAD_MESSAGE_COMPLETED("thread.message.completed")
- ThreadMessageIncomplete
  - Message data
    
    Represents a message within a thread.
  - JsonValue; event "thread.message.incomplete"constant
    - THREAD_MESSAGE_INCOMPLETE("thread.message.incomplete")
- ErrorEvent
  - ErrorObject data
    - Optional<String> code
    - String message
    - Optional<String> param
    - String type
  - JsonValue; event "error"constant
    - ERROR("error")

Assistant Tool

class AssistantTool:

Code Interpreter Tool

class CodeInterpreterTool:
- JsonValue; type "code_interpreter"constant
  
  The type of tool being defined: code_interpreter
  - CODE_INTERPRETER("code_interpreter")

File Search Tool

class FileSearchTool:
- JsonValue; type "file_search"constant
  
  The type of tool being defined: file_search
  - FILE_SEARCH("file_search")
- Optional<FileSearch> fileSearch
  
  Overrides for the file search tool.
  - Optional<Long> maxNumResults
    
    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.
  - Optional<RankingOptions> rankingOptions
    
    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.
    - double scoreThreshold
      
      The score threshold for the file search. All values must be a floating point number between 0 and 1.
    - Optional<Ranker> ranker
      
      The ranker to use for the file search. If not specified will use the auto ranker.
      - AUTO("auto")
      - DEFAULT_2024_08_21("default_2024_08_21")

Function Tool

class FunctionTool:
- FunctionDefinition function
  - String name
    
    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.
  - Optional<String> description
    
    A description of what the function does, used by the model to choose when and how to call the function.
  - Optional<FunctionParameters> parameters
    
    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.
  - Optional<Boolean> strict
    
    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.
- JsonValue; type "function"constant
  
  The type of tool being defined: function
  - FUNCTION("function")

Message Stream Event

class MessageStreamEvent: A class that can be one of several variants.union

Occurs when a message is created.
- ThreadMessageCreated
  - Message data
    
    Represents a message within a thread.
    - String id
      
      The identifier, which can be referenced in API endpoints.
    - Optional<String> assistantId
      
      If applicable, the ID of the assistant that authored this message.
    - Optional<List<Attachment>> attachments
      
      A list of files attached to the message, and the tools they were added to.
      - Optional<String> fileId
        
        The ID of the file to attach to the message.
      - Optional<List<Tool>> tools
        
        The tools to add this file to.
        
        class CodeInterpreterTool:
        
        JsonValue; type "code_interpreter"constant
        
        The type of tool being defined: code_interpreter
        
        CODE_INTERPRETER("code_interpreter")
        
        JsonValue;
        
        JsonValue; type "file_search"constant
        
        The type of tool being defined: file_search
        
        FILE_SEARCH("file_search")
    - Optional<Long> completedAt
      
      The Unix timestamp (in seconds) for when the message was completed.
    - List<JsonValue> content
      
      The content of the message in array of text and/or images.
    - long createdAt
      
      The Unix timestamp (in seconds) for when the message was created.
    - Optional<Long> incompleteAt
      
      The Unix timestamp (in seconds) for when the message was marked as incomplete.
    - Optional<IncompleteDetails> incompleteDetails
      
      On an incomplete message, details about why the message is incomplete.
      - Reason reason
        
        The reason the message is incomplete.
        
        CONTENT_FILTER("content_filter")
        
        MAX_TOKENS("max_tokens")
        
        RUN_CANCELLED("run_cancelled")
        
        RUN_EXPIRED("run_expired")
        
        RUN_FAILED("run_failed")
    - Optional<Metadata> metadata
      
      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.
    - JsonValue; object_ "thread.message"constant
      
      The object type, which is always thread.message.
      - THREAD_MESSAGE("thread.message")
    - Role role
      
      The entity that produced the message. One of user or assistant.
      - USER("user")
      - ASSISTANT("assistant")
    - Optional<String> runId
      
      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 status
      
      The status of the message, which can be either in_progress, incomplete, or completed.
      - IN_PROGRESS("in_progress")
      - INCOMPLETE("incomplete")
      - COMPLETED("completed")
    - String threadId
      
      The thread ID that this message belongs to.
  - JsonValue; event "thread.message.created"constant
    - THREAD_MESSAGE_CREATED("thread.message.created")
- ThreadMessageInProgress
  - Message data
    
    Represents a message within a thread.
  - JsonValue; event "thread.message.in_progress"constant
    - THREAD_MESSAGE_IN_PROGRESS("thread.message.in_progress")
- ThreadMessageDelta
  - MessageDeltaEvent data
    
    Represents a message delta i.e. any changed fields on a message during streaming.
    - String id
      
      The identifier of the message, which can be referenced in API endpoints.
    - MessageDelta delta
      
      The delta containing the fields that have changed on the Message.
      - Optional<List<JsonValue>> content
        
        The content of the message in array of text and/or images.
      - Optional<Role> role
        
        The entity that produced the message. One of user or assistant.
        
        USER("user")
        
        ASSISTANT("assistant")
    - JsonValue; object_ "thread.message.delta"constant
      
      The object type, which is always thread.message.delta.
      - THREAD_MESSAGE_DELTA("thread.message.delta")
  - JsonValue; event "thread.message.delta"constant
    - THREAD_MESSAGE_DELTA("thread.message.delta")
- ThreadMessageCompleted
  - Message data
    
    Represents a message within a thread.
  - JsonValue; event "thread.message.completed"constant
    - THREAD_MESSAGE_COMPLETED("thread.message.completed")
- ThreadMessageIncomplete
  - Message data
    
    Represents a message within a thread.
  - JsonValue; event "thread.message.incomplete"constant
    - THREAD_MESSAGE_INCOMPLETE("thread.message.incomplete")

Run Step Stream Event

class RunStepStreamEvent: A class that can be one of several variants.union

Occurs when a run step is created.
- ThreadRunStepCreated
  - RunStep data
    
    Represents a step in execution of a run.
    - String id
      
      The identifier of the run step, which can be referenced in API endpoints.
    - String assistantId
      
      The ID of the assistant associated with the run step.
    - Optional<Long> cancelledAt
      
      The Unix timestamp (in seconds) for when the run step was cancelled.
    - Optional<Long> completedAt
      
      The Unix timestamp (in seconds) for when the run step completed.
    - long createdAt
      
      The Unix timestamp (in seconds) for when the run step was created.
    - Optional<Long> expiredAt
      
      The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.
    - Optional<Long> failedAt
      
      The Unix timestamp (in seconds) for when the run step failed.
    - Optional<LastError> lastError
      
      The last error associated with this run step. Will be null if there are no errors.
      - Code code
        
        One of server_error or rate_limit_exceeded.
        
        SERVER_ERROR("server_error")
        
        RATE_LIMIT_EXCEEDED("rate_limit_exceeded")
      - String message
        
        A human-readable description of the error.
    - Optional<Metadata> metadata
      
      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.
    - JsonValue; object_ "thread.run.step"constant
      
      The object type, which is always thread.run.step.
      - THREAD_RUN_STEP("thread.run.step")
    - String runId
      
      The ID of the run that this run step is a part of.
    - Status status
      
      The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.
      - IN_PROGRESS("in_progress")
      - CANCELLED("cancelled")
      - FAILED("failed")
      - COMPLETED("completed")
      - EXPIRED("expired")
    - StepDetails stepDetails
      
      The details of the run step.
      - class MessageCreationStepDetails:
        
        Details of the message creation by the run step.
        
        MessageCreation messageCreation
        
        String messageId
        
        The ID of the message that was created by this run step.
        
        JsonValue; type "message_creation"constant
        
        Always message_creation.
        
        MESSAGE_CREATION("message_creation")
      - class ToolCallsStepDetails:
        
        Details of the tool call.
        
        List<JsonValue> toolCalls
        
        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.
        
        JsonValue; type "tool_calls"constant
        
        Always tool_calls.
        
        TOOL_CALLS("tool_calls")
    - String threadId
      
      The ID of the thread that was run.
    - Type type
      
      The type of run step, which can be either message_creation or tool_calls.
      - MESSAGE_CREATION("message_creation")
      - TOOL_CALLS("tool_calls")
    - Optional<Usage> usage
      
      Usage statistics related to the run step. This value will be null while the run step's status is in_progress.
      - long completionTokens
        
        Number of completion tokens used over the course of the run step.
      - long promptTokens
        
        Number of prompt tokens used over the course of the run step.
      - long totalTokens
        
        Total number of tokens used (prompt + completion).
  - JsonValue; event "thread.run.step.created"constant
    - THREAD_RUN_STEP_CREATED("thread.run.step.created")
- ThreadRunStepInProgress
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.in_progress"constant
    - THREAD_RUN_STEP_IN_PROGRESS("thread.run.step.in_progress")
- ThreadRunStepDelta
  - RunStepDeltaEvent data
    
    Represents a run step delta i.e. any changed fields on a run step during streaming.
    - String id
      
      The identifier of the run step, which can be referenced in API endpoints.
    - JsonValue delta
    - JsonValue; object_ "thread.run.step.delta"constant
      
      The object type, which is always thread.run.step.delta.
      - THREAD_RUN_STEP_DELTA("thread.run.step.delta")
  - JsonValue; event "thread.run.step.delta"constant
    - THREAD_RUN_STEP_DELTA("thread.run.step.delta")
- ThreadRunStepCompleted
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.completed"constant
    - THREAD_RUN_STEP_COMPLETED("thread.run.step.completed")
- ThreadRunStepFailed
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.failed"constant
    - THREAD_RUN_STEP_FAILED("thread.run.step.failed")
- ThreadRunStepCancelled
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.cancelled"constant
    - THREAD_RUN_STEP_CANCELLED("thread.run.step.cancelled")
- ThreadRunStepExpired
  - RunStep data
    
    Represents a step in execution of a run.
  - JsonValue; event "thread.run.step.expired"constant
    - THREAD_RUN_STEP_EXPIRED("thread.run.step.expired")

Run Stream Event

class RunStreamEvent: A class that can be one of several variants.union

Occurs when a new run is created.
- ThreadRunCreated
  - Run data
    
    Represents an execution run on a thread.
    - String id
      
      The identifier, which can be referenced in API endpoints.
    - String assistantId
      
      The ID of the assistant used for execution of this run.
    - Optional<Long> cancelledAt
      
      The Unix timestamp (in seconds) for when the run was cancelled.
    - Optional<Long> completedAt
      
      The Unix timestamp (in seconds) for when the run was completed.
    - long createdAt
      
      The Unix timestamp (in seconds) for when the run was created.
    - Optional<Long> expiresAt
      
      The Unix timestamp (in seconds) for when the run will expire.
    - Optional<Long> failedAt
      
      The Unix timestamp (in seconds) for when the run failed.
    - Optional<IncompleteDetails> incompleteDetails
      
      Details on why the run is incomplete. Will be null if the run is not incomplete.
      - Optional<Reason> reason
        
        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_completion_tokens")
        
        MAX_PROMPT_TOKENS("max_prompt_tokens")
    - String instructions
      
      The instructions that the assistant used for this run.
    - Optional<LastError> lastError
      
      The last error associated with this run. Will be null if there are no errors.
      - Code code
        
        One of server_error, rate_limit_exceeded, or invalid_prompt.
        
        SERVER_ERROR("server_error")
        
        RATE_LIMIT_EXCEEDED("rate_limit_exceeded")
        
        INVALID_PROMPT("invalid_prompt")
      - String message
        
        A human-readable description of the error.
    - Optional<Long> maxCompletionTokens
      
      The maximum number of completion tokens specified to have been used over the course of the run.
    - Optional<Long> maxPromptTokens
      
      The maximum number of prompt tokens specified to have been used over the course of the run.
    - Optional<Metadata> metadata
      
      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.
    - String model
      
      The model that the assistant used for this run.
    - JsonValue; object_ "thread.run"constant
      
      The object type, which is always thread.run.
      - THREAD_RUN("thread.run")
    - boolean parallelToolCalls
      
      Whether to enable parallel function calling during tool use.
    - Optional<RequiredAction> requiredAction
      
      Details on the action required to continue the run. Will be null if no action is required.
      - SubmitToolOutputs submitToolOutputs
        
        Details on the tool outputs needed for this run to continue.
        
        List<RequiredActionFunctionToolCall> toolCalls
        
        A list of the relevant tool calls.
        
        String id
        
        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 function
        
        The function definition.
        
        String arguments
        
        The arguments that the model expects you to pass to the function.
        
        String name
        
        The name of the function.
        
        JsonValue; type "function"constant
        
        The type of tool call the output is required for. For now, this is always function.
        
        FUNCTION("function")
      - JsonValue; type "submit_tool_outputs"constant
        
        For now, this is always submit_tool_outputs.
        
        SUBMIT_TOOL_OUTPUTS("submit_tool_outputs")
    - Optional<AssistantResponseFormatOption> responseFormat
      
      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.
      - JsonValue;
        
        AUTO("auto")
      - class ResponseFormatText:
        
        Default response format. Used to generate text responses.
        
        JsonValue; type "text"constant
        
        The type of response format being defined. Always text.
        
        TEXT("text")
      - class ResponseFormatJsonObject:
        
        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.
        
        JsonValue; type "json_object"constant
        
        The type of response format being defined. Always json_object.
        
        JSON_OBJECT("json_object")
      - class ResponseFormatJsonSchema:
        
        JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
        
        JsonSchema jsonSchema
        
        Structured Outputs configuration options, including a JSON Schema.
        
        String name
        
        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.
        
        Optional<String> description
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
        
        Optional<Schema> schema
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
        
        Optional<Boolean> strict
        
        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.
        
        JsonValue; type "json_schema"constant
        
        The type of response format being defined. Always json_schema.
        
        JSON_SCHEMA("json_schema")
    - Optional<Long> startedAt
      
      The Unix timestamp (in seconds) for when the run was started.
    - JsonValue status
    - String threadId
      
      The ID of the thread that was executed on as a part of this run.
    - Optional<AssistantToolChoiceOption> toolChoice
      
      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.
      - Auto
        
        NONE("none")
        
        AUTO("auto")
        
        REQUIRED("required")
      - class AssistantToolChoice:
        
        Specifies a tool the model should use. Use to force the model to call a specific tool.
        
        Type type
        
        The type of the tool. If type is function, the function name must be set
        
        FUNCTION("function")
        
        CODE_INTERPRETER("code_interpreter")
        
        FILE_SEARCH("file_search")
        
        Optional<AssistantToolChoiceFunction> function
        
        String name
        
        The name of the function to call.
    - List<JsonValue> tools
      
      The list of tools that the assistant used for this run.
    - Optional<TruncationStrategy> truncationStrategy
      
      Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.
      - Type type
        
        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("auto")
        
        LAST_MESSAGES("last_messages")
      - Optional<Long> lastMessages
        
        The number of most recent messages from the thread when constructing the context for the run.
    - Optional<Usage> usage
      
      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.).
      - long completionTokens
        
        Number of completion tokens used over the course of the run.
      - long promptTokens
        
        Number of prompt tokens used over the course of the run.
      - long totalTokens
        
        Total number of tokens used (prompt + completion).
    - Optional<Double> temperature
      
      The sampling temperature used for this run. If not set, defaults to 1.
    - Optional<Double> topP
      
      The nucleus sampling value used for this run. If not set, defaults to 1.
  - JsonValue; event "thread.run.created"constant
    - THREAD_RUN_CREATED("thread.run.created")
- ThreadRunQueued
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.queued"constant
    - THREAD_RUN_QUEUED("thread.run.queued")
- ThreadRunInProgress
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.in_progress"constant
    - THREAD_RUN_IN_PROGRESS("thread.run.in_progress")
- ThreadRunRequiresAction
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.requires_action"constant
    - THREAD_RUN_REQUIRES_ACTION("thread.run.requires_action")
- ThreadRunCompleted
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.completed"constant
    - THREAD_RUN_COMPLETED("thread.run.completed")
- ThreadRunIncomplete
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.incomplete"constant
    - THREAD_RUN_INCOMPLETE("thread.run.incomplete")
- ThreadRunFailed
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.failed"constant
    - THREAD_RUN_FAILED("thread.run.failed")
- ThreadRunCancelling
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.cancelling"constant
    - THREAD_RUN_CANCELLING("thread.run.cancelling")
- ThreadRunCancelled
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.cancelled"constant
    - THREAD_RUN_CANCELLED("thread.run.cancelled")
- ThreadRunExpired
  - Run data
    
    Represents an execution run on a thread.
  - JsonValue; event "thread.run.expired"constant
    - THREAD_RUN_EXPIRED("thread.run.expired")

Thread Stream Event

class ThreadStreamEvent:

Occurs when a new thread is created.
- Thread data
  
  Represents a thread that contains messages.
  - String id
    
    The identifier, which can be referenced in API endpoints.
  - long createdAt
    
    The Unix timestamp (in seconds) for when the thread was created.
  - Optional<Metadata> metadata
    
    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.
  - JsonValue; object_ "thread"constant
    
    The object type, which is always thread.
    - THREAD("thread")
  - Optional<ToolResources> toolResources
    
    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.
    - Optional<CodeInterpreter> codeInterpreter
      - Optional<List<String>> fileIds
        
        A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
    - Optional<FileSearch> fileSearch
      - Optional<List<String>> vectorStoreIds
        
        The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.
- JsonValue; event "thread.created"constant
  - THREAD_CREATED("thread.created")
- Optional<Boolean> enabled
  
  Whether to enable input audio transcription.

java/resources/beta/subresources/assistants/index.md +442 −2382

6 6

7**get** `/assistants`7**get** `/assistants`

8 8

~~9Returns a list of assistants.~~9List assistants

10 10

11### Parameters11### Parameters

12 12

77 77

78 - `ASSISTANT("assistant")`78 - `ASSISTANT("assistant")`

79 79

~~80 - `List<AssistantTool> tools`~~80 - `List<JsonValue> tools`

81 81

82 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`.82 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`.

83 83

~~84 - `class CodeInterpreterTool:`~~

~~86 - `JsonValue; type "code_interpreter"constant`~~

~~88 The type of tool being defined: `code_interpreter`~~

~~90 - `CODE_INTERPRETER("code_interpreter")`~~

~~92 - `class FileSearchTool:`~~

~~94 - `JsonValue; type "file_search"constant`~~

~~96 The type of tool being defined: `file_search`~~

~~98 - `FILE_SEARCH("file_search")`~~

100 - `Optional<FileSearch> fileSearch`

~~101~~

102 Overrides for the file search tool.

~~103~~

104 - `Optional<Long> maxNumResults`

~~105~~

106 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.

~~107~~

108 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~109~~

110 - `Optional<RankingOptions> rankingOptions`

~~111~~

112 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.

~~113~~

114 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~115~~

116 - `double scoreThreshold`

~~117~~

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

~~119~~

120 - `Optional<Ranker> ranker`

~~121~~

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

~~123~~

124 - `AUTO("auto")`

~~125~~

126 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~127~~

128 - `class FunctionTool:`

~~129~~

130 - `FunctionDefinition function`

~~131~~

132 - `String name`

~~133~~

134 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.

~~135~~

136 - `Optional<String> description`

~~137~~

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

~~139~~

140 - `Optional<FunctionParameters> parameters`

~~141~~

142 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~143~~

144 Omitting `parameters` defines a function with an empty parameter list.

~~145~~

146 - `Optional<Boolean> strict`

~~147~~

148 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](https://platform.openai.com/docs/guides/function-calling).

~~149~~

150 - `JsonValue; type "function"constant`

~~151~~

152 The type of tool being defined: `function`

~~153~~

154 - `FUNCTION("function")`

~~155~~

156 - `Optional<AssistantResponseFormatOption> responseFormat`84 - `Optional<AssistantResponseFormatOption> responseFormat`

157 85

158 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.86 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.

292 "name": "name",220 "name": "name",

293 "object": "assistant",221 "object": "assistant",

294 "tools": [222 "tools": [

295 {223 {}

296 "type": "code_interpreter"

297 }

298 ],224 ],

299 "response_format": "auto",225 "response_format": "auto",

300 "temperature": 1,226 "temperature": 1,

326 252

327**post** `/assistants`253**post** `/assistants`

328 254

329Create an assistant with a model and instructions.255Create assistant

330 256

331### Parameters257### Parameters

332 258

359 285

360 - `Optional<ReasoningEffort> reasoningEffort`286 - `Optional<ReasoningEffort> reasoningEffort`

361 287

362 Constrains effort on reasoning for288 Constrains effort on reasoning for reasoning models. Currently supported

363 [reasoning models](https://platform.openai.com/docs/guides/reasoning).289 values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`.

364 Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, and `xhigh`. Reducing290 Reducing reasoning effort can result in faster responses and fewer tokens

365 reasoning effort can result in faster responses and fewer tokens used291 used on reasoning in a response. Not all reasoning models support every

366 on reasoning in a response.292 value. See the

~~367~~ 293 [reasoning guide](https://platform.openai.com/docs/guides/reasoning)

368 - `gpt-5.1` defaults to `none`, which does not perform reasoning. The supported reasoning values for `gpt-5.1` are `none`, `low`, `medium`, and `high`. Tool calls are supported for all reasoning values in gpt-5.1.294 for model-specific support.

369 - All models before `gpt-5.1` default to `medium` reasoning effort, and do not support `none`.

370 - The `gpt-5-pro` model defaults to (and only supports) `high` reasoning effort.

371 - `xhigh` is supported for all models after `gpt-5.1-codex-max`.

372 295

373 - `Optional<AssistantResponseFormatOption> responseFormat`296 - `Optional<AssistantResponseFormatOption> responseFormat`

374 297

449 Keys are strings with a maximum length of 64 characters. Values are strings372 Keys are strings with a maximum length of 64 characters. Values are strings

450 with a maximum length of 512 characters.373 with a maximum length of 512 characters.

451 374

452 - `Optional<List<AssistantTool>> tools`375 - `Optional<List<JsonValue>> tools`

453 376

454 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`.377 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`.

455 378

456 - `class CodeInterpreterTool:`

~~457~~

458 - `JsonValue; type "code_interpreter"constant`

~~459~~

460 The type of tool being defined: `code_interpreter`

~~461~~

462 - `CODE_INTERPRETER("code_interpreter")`

~~463~~

464 - `class FileSearchTool:`

~~465~~

466 - `JsonValue; type "file_search"constant`

~~467~~

468 The type of tool being defined: `file_search`

~~469~~

470 - `FILE_SEARCH("file_search")`

~~471~~

472 - `Optional<FileSearch> fileSearch`

~~473~~

474 Overrides for the file search tool.

~~475~~

476 - `Optional<Long> maxNumResults`

~~477~~

478 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.

~~479~~

480 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~481~~

482 - `Optional<RankingOptions> rankingOptions`

~~483~~

484 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.

~~485~~

486 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~487~~

488 - `double scoreThreshold`

~~489~~

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

~~491~~

492 - `Optional<Ranker> ranker`

~~493~~

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

~~495~~

496 - `AUTO("auto")`

~~497~~

498 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~499~~

500 - `class FunctionTool:`

~~501~~

502 - `FunctionDefinition function`

~~503~~

504 - `String name`

~~505~~

506 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.

~~507~~

508 - `Optional<String> description`

~~509~~

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

~~511~~

512 - `Optional<FunctionParameters> parameters`

~~513~~

514 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~515~~

516 Omitting `parameters` defines a function with an empty parameter list.

~~517~~

518 - `Optional<Boolean> strict`

~~519~~

520 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](https://platform.openai.com/docs/guides/function-calling).

~~521~~

522 - `JsonValue; type "function"constant`

~~523~~

524 The type of tool being defined: `function`

~~525~~

526 - `FUNCTION("function")`

~~527~~

528 - `Optional<Double> topP`379 - `Optional<Double> topP`

529 380

530 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.381 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.

576 427

577 - `ASSISTANT("assistant")`428 - `ASSISTANT("assistant")`

578 429

579 - `List<AssistantTool> tools`430 - `List<JsonValue> tools`

580 431

581 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`.432 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`.

582 433

583 - `class CodeInterpreterTool:`

~~584~~

585 - `JsonValue; type "code_interpreter"constant`

~~586~~

587 The type of tool being defined: `code_interpreter`

~~588~~

589 - `CODE_INTERPRETER("code_interpreter")`

~~590~~

591 - `class FileSearchTool:`

~~592~~

593 - `JsonValue; type "file_search"constant`

~~594~~

595 The type of tool being defined: `file_search`

~~596~~

597 - `FILE_SEARCH("file_search")`

~~598~~

599 - `Optional<FileSearch> fileSearch`

~~600~~

601 Overrides for the file search tool.

~~602~~

603 - `Optional<Long> maxNumResults`

~~604~~

605 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.

~~606~~

607 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~608~~

609 - `Optional<RankingOptions> rankingOptions`

~~610~~

611 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.

~~612~~

613 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~614~~

615 - `double scoreThreshold`

~~616~~

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

~~618~~

619 - `Optional<Ranker> ranker`

~~620~~

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

~~622~~

623 - `AUTO("auto")`

~~624~~

625 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~626~~

627 - `class FunctionTool:`

~~628~~

629 - `FunctionDefinition function`

~~630~~

631 - `String name`

~~632~~

633 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.

~~634~~

635 - `Optional<String> description`

~~636~~

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

~~638~~

639 - `Optional<FunctionParameters> parameters`

~~640~~

641 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~642~~

643 Omitting `parameters` defines a function with an empty parameter list.

~~644~~

645 - `Optional<Boolean> strict`

~~646~~

647 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](https://platform.openai.com/docs/guides/function-calling).

~~648~~

649 - `JsonValue; type "function"constant`

~~650~~

651 The type of tool being defined: `function`

~~652~~

653 - `FUNCTION("function")`

~~654~~

655 - `Optional<AssistantResponseFormatOption> responseFormat`434 - `Optional<AssistantResponseFormatOption> responseFormat`

656 435

657 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.436 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.

793 "name": "name",572 "name": "name",

794 "object": "assistant",573 "object": "assistant",

795 "tools": [574 "tools": [

796 {575 {}

797 "type": "code_interpreter"

798 }

799 ],576 ],

800 "response_format": "auto",577 "response_format": "auto",

801 "temperature": 1,578 "temperature": 1,

821 598

822**get** `/assistants/{assistant_id}`599**get** `/assistants/{assistant_id}`

823 600

824Retrieves an assistant.601Retrieve assistant

825 602

826### Parameters603### Parameters

827 604

874 651

875 - `ASSISTANT("assistant")`652 - `ASSISTANT("assistant")`

876 653

877 - `List<AssistantTool> tools`654 - `List<JsonValue> tools`

878 655

879 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`.656 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`.

880 657

881 - `class CodeInterpreterTool:`658 - `Optional<AssistantResponseFormatOption> responseFormat`

882 659

883 - `JsonValue; type "code_interpreter"constant`660 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.

884 661

885 The type of tool being defined: `code_interpreter`662 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](https://platform.openai.com/docs/guides/structured-outputs).

886 663

887 - `CODE_INTERPRETER("code_interpreter")`664 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

888 665

889 - `class FileSearchTool:`666 **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.

890 667

891 - `JsonValue; type "file_search"constant`668 - `JsonValue;`

892 669

893 The type of tool being defined: `file_search`670 - `AUTO("auto")`

894 671

895 - `FILE_SEARCH("file_search")`672 - `class ResponseFormatText:`

896 673

897 - `Optional<FileSearch> fileSearch`674 Default response format. Used to generate text responses.

898 675

899 Overrides for the file search tool.676 - `JsonValue; type "text"constant`

900 677

901 - `Optional<Long> maxNumResults`678 The type of response format being defined. Always `text`.

902 679

903 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.680 - `TEXT("text")`

904 681

905 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.682 - `class ResponseFormatJsonObject:`

906 683

907 - `Optional<RankingOptions> rankingOptions`684 JSON object response format. An older method of generating JSON responses.

685 Using `json_schema` is recommended for models that support it. Note that the

686 model will not generate JSON without a system or user message instructing it

687 to do so.

908 688

909 The ranking options for the file search. If not specified, the file search tool will use the `auto` ranker and a score_threshold of 0.689 - `JsonValue; type "json_object"constant`

910 690

911 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.691 The type of response format being defined. Always `json_object`.

912 692

913 - `double scoreThreshold`693 - `JSON_OBJECT("json_object")`

914 694

915 The score threshold for the file search. All values must be a floating point number between 0 and 1.695 - `class ResponseFormatJsonSchema:`

916 696

917 - `Optional<Ranker> ranker`697 JSON Schema response format. Used to generate structured JSON responses.

~~918~~ 698 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).

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

~~920~~

921 - `AUTO("auto")`

~~922~~

923 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~924~~

925 - `class FunctionTool:`

~~926~~

927 - `FunctionDefinition function`

~~928~~

929 - `String name`

~~930~~

931 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.

~~932~~

933 - `Optional<String> description`

~~934~~

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

~~936~~

937 - `Optional<FunctionParameters> parameters`

~~938~~

939 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~940~~

941 Omitting `parameters` defines a function with an empty parameter list.

~~942~~

943 - `Optional<Boolean> strict`

~~944~~

945 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](https://platform.openai.com/docs/guides/function-calling).

~~946~~

947 - `JsonValue; type "function"constant`

~~948~~

949 The type of tool being defined: `function`

~~950~~

951 - `FUNCTION("function")`

~~952~~

953 - `Optional<AssistantResponseFormatOption> responseFormat`

~~954~~

955 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.

~~956~~

957 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](https://platform.openai.com/docs/guides/structured-outputs).

~~958~~

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

~~960~~

961 **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.

~~962~~

963 - `JsonValue;`

~~964~~

965 - `AUTO("auto")`

~~966~~

967 - `class ResponseFormatText:`

~~968~~

969 Default response format. Used to generate text responses.

~~970~~

971 - `JsonValue; type "text"constant`

~~972~~

973 The type of response format being defined. Always `text`.

~~974~~

975 - `TEXT("text")`

~~976~~

977 - `class ResponseFormatJsonObject:`

~~978~~

979 JSON object response format. An older method of generating JSON responses.

980 Using `json_schema` is recommended for models that support it. Note that the

981 model will not generate JSON without a system or user message instructing it

982 to do so.

~~983~~

984 - `JsonValue; type "json_object"constant`

~~985~~

986 The type of response format being defined. Always `json_object`.

~~987~~

988 - `JSON_OBJECT("json_object")`

~~989~~

990 - `class ResponseFormatJsonSchema:`

~~991~~

992 JSON Schema response format. Used to generate structured JSON responses.

993 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).

994 699

995 - `JsonSchema jsonSchema`700 - `JsonSchema jsonSchema`

996 701

1087 "name": "name",792 "name": "name",

1088 "object": "assistant",793 "object": "assistant",

1089 "tools": [794 "tools": [

1090 {795 {}

1091 "type": "code_interpreter"

1092 }

1093 ],796 ],

1094 "response_format": "auto",797 "response_format": "auto",

1095 "temperature": 1,798 "temperature": 1,

1115 818

1116**post** `/assistants/{assistant_id}`819**post** `/assistants/{assistant_id}`

1117 820

1118Modifies an assistant.821Modify assistant

1119 822

1120### Parameters823### Parameters

1121 824

1234 937

1235 - `Optional<ReasoningEffort> reasoningEffort`938 - `Optional<ReasoningEffort> reasoningEffort`

1236 939

1237 Constrains effort on reasoning for940 Constrains effort on reasoning for reasoning models. Currently supported

1238 [reasoning models](https://platform.openai.com/docs/guides/reasoning).941 values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`.

1239 Currently supported values are `none`, `minimal`, `low`, `medium`, `high`, and `xhigh`. Reducing942 Reducing reasoning effort can result in faster responses and fewer tokens

1240 reasoning effort can result in faster responses and fewer tokens used943 used on reasoning in a response. Not all reasoning models support every

1241 on reasoning in a response.944 value. See the

~~1242~~ 945 [reasoning guide](https://platform.openai.com/docs/guides/reasoning)

1243 - `gpt-5.1` defaults to `none`, which does not perform reasoning. The supported reasoning values for `gpt-5.1` are `none`, `low`, `medium`, and `high`. Tool calls are supported for all reasoning values in gpt-5.1.946 for model-specific support.

1244 - All models before `gpt-5.1` default to `medium` reasoning effort, and do not support `none`.

1245 - The `gpt-5-pro` model defaults to (and only supports) `high` reasoning effort.

1246 - `xhigh` is supported for all models after `gpt-5.1-codex-max`.

1247 947

1248 - `Optional<AssistantResponseFormatOption> responseFormat`948 - `Optional<AssistantResponseFormatOption> responseFormat`

1249 949

1275 975

1276 Overrides the [vector store](https://platform.openai.com/docs/api-reference/vector-stores/object) attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.976 Overrides the [vector store](https://platform.openai.com/docs/api-reference/vector-stores/object) attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.

1277 977

1278 - `Optional<List<AssistantTool>> tools`978 - `Optional<List<JsonValue>> tools`

1279 979

1280 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`.980 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`.

1281 981

1282 - `class CodeInterpreterTool:`

~~1283~~

1284 - `JsonValue; type "code_interpreter"constant`

~~1285~~

1286 The type of tool being defined: `code_interpreter`

~~1287~~

1288 - `CODE_INTERPRETER("code_interpreter")`

~~1289~~

1290 - `class FileSearchTool:`

~~1291~~

1292 - `JsonValue; type "file_search"constant`

~~1293~~

1294 The type of tool being defined: `file_search`

~~1295~~

1296 - `FILE_SEARCH("file_search")`

~~1297~~

1298 - `Optional<FileSearch> fileSearch`

~~1299~~

1300 Overrides for the file search tool.

~~1301~~

1302 - `Optional<Long> maxNumResults`

~~1303~~

1304 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.

~~1305~~

1306 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~1307~~

1308 - `Optional<RankingOptions> rankingOptions`

~~1309~~

1310 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.

~~1311~~

1312 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~1313~~

1314 - `double scoreThreshold`

~~1315~~

1316 The score threshold for the file search. All values must be a floating point number between 0 and 1.

~~1317~~

1318 - `Optional<Ranker> ranker`

~~1319~~

1320 The ranker to use for the file search. If not specified will use the `auto` ranker.

~~1321~~

1322 - `AUTO("auto")`

~~1323~~

1324 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~1325~~

1326 - `class FunctionTool:`

~~1327~~

1328 - `FunctionDefinition function`

~~1329~~

1330 - `String name`

~~1331~~

1332 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.

~~1333~~

1334 - `Optional<String> description`

~~1335~~

1336 A description of what the function does, used by the model to choose when and how to call the function.

~~1337~~

1338 - `Optional<FunctionParameters> parameters`

~~1339~~

1340 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~1341~~

1342 Omitting `parameters` defines a function with an empty parameter list.

~~1343~~

1344 - `Optional<Boolean> strict`

~~1345~~

1346 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](https://platform.openai.com/docs/guides/function-calling).

~~1347~~

1348 - `JsonValue; type "function"constant`

~~1349~~

1350 The type of tool being defined: `function`

~~1351~~

1352 - `FUNCTION("function")`

~~1353~~

1354 - `Optional<Double> topP`982 - `Optional<Double> topP`

1355 983

1356 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.984 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.

1402 1030

1403 - `ASSISTANT("assistant")`1031 - `ASSISTANT("assistant")`

1404 1032

1405 - `List<AssistantTool> tools`1033 - `List<JsonValue> tools`

1406 1034

1407 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`.1035 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`.

1408 1036

1409 - `class CodeInterpreterTool:`

~~1410~~

1411 - `JsonValue; type "code_interpreter"constant`

~~1412~~

1413 The type of tool being defined: `code_interpreter`

~~1414~~

1415 - `CODE_INTERPRETER("code_interpreter")`

~~1416~~

1417 - `class FileSearchTool:`

~~1418~~

1419 - `JsonValue; type "file_search"constant`

~~1420~~

1421 The type of tool being defined: `file_search`

~~1422~~

1423 - `FILE_SEARCH("file_search")`

~~1424~~

1425 - `Optional<FileSearch> fileSearch`

~~1426~~

1427 Overrides for the file search tool.

~~1428~~

1429 - `Optional<Long> maxNumResults`

~~1430~~

1431 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.

~~1432~~

1433 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~1434~~

1435 - `Optional<RankingOptions> rankingOptions`

~~1436~~

1437 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.

~~1438~~

1439 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~1440~~

1441 - `double scoreThreshold`

~~1442~~

1443 The score threshold for the file search. All values must be a floating point number between 0 and 1.

~~1444~~

1445 - `Optional<Ranker> ranker`

~~1446~~

1447 The ranker to use for the file search. If not specified will use the `auto` ranker.

~~1448~~

1449 - `AUTO("auto")`

~~1450~~

1451 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~1452~~

1453 - `class FunctionTool:`

~~1454~~

1455 - `FunctionDefinition function`

~~1456~~

1457 - `String name`

~~1458~~

1459 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.

~~1460~~

1461 - `Optional<String> description`

~~1462~~

1463 A description of what the function does, used by the model to choose when and how to call the function.

~~1464~~

1465 - `Optional<FunctionParameters> parameters`

~~1466~~

1467 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~1468~~

1469 Omitting `parameters` defines a function with an empty parameter list.

~~1470~~

1471 - `Optional<Boolean> strict`

~~1472~~

1473 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](https://platform.openai.com/docs/guides/function-calling).

~~1474~~

1475 - `JsonValue; type "function"constant`

~~1476~~

1477 The type of tool being defined: `function`

~~1478~~

1479 - `FUNCTION("function")`

~~1480~~

1481 - `Optional<AssistantResponseFormatOption> responseFormat`1037 - `Optional<AssistantResponseFormatOption> responseFormat`

1482 1038

1483 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.1039 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.

1615 "name": "name",1171 "name": "name",

1616 "object": "assistant",1172 "object": "assistant",

1617 "tools": [1173 "tools": [

1618 {1174 {}

1619 "type": "code_interpreter"

1620 }

1621 ],1175 ],

1622 "response_format": "auto",1176 "response_format": "auto",

1623 "temperature": 1,1177 "temperature": 1,

1643 1197

1644**delete** `/assistants/{assistant_id}`1198**delete** `/assistants/{assistant_id}`

1645 1199

1646Delete an assistant.1200Delete assistant

1647 1201

1648### Parameters1202### Parameters

1649 1203

1741 1295

1742 - `ASSISTANT("assistant")`1296 - `ASSISTANT("assistant")`

1743 1297

1744 - `List<AssistantTool> tools`1298 - `List<JsonValue> tools`

1745 1299

1746 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`.1300 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`.

1747 1301

1748 - `class CodeInterpreterTool:`1302 - `Optional<AssistantResponseFormatOption> responseFormat`

1749 1303

1750 - `JsonValue; type "code_interpreter"constant`1304 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.

1751 1305

1752 The type of tool being defined: `code_interpreter`1306 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](https://platform.openai.com/docs/guides/structured-outputs).

1753 1307

1754 - `CODE_INTERPRETER("code_interpreter")`1308 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

1755 1309

1756 - `class FileSearchTool:`1310 **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.

1757 1311

1758 - `JsonValue; type "file_search"constant`1312 - `JsonValue;`

1759 1313

1760 The type of tool being defined: `file_search`1314 - `AUTO("auto")`

1761 1315

1762 - `FILE_SEARCH("file_search")`1316 - `class ResponseFormatText:`

1763 1317

1764 - `Optional<FileSearch> fileSearch`1318 Default response format. Used to generate text responses.

1765 1319

1766 Overrides for the file search tool.1320 - `JsonValue; type "text"constant`

1767 1321

1768 - `Optional<Long> maxNumResults`1322 The type of response format being defined. Always `text`.

1769 1323

1770 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.1324 - `TEXT("text")`

1771 1325

1772 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.1326 - `class ResponseFormatJsonObject:`

1773 1327

1774 - `Optional<RankingOptions> rankingOptions`1328 JSON object response format. An older method of generating JSON responses.

1329 Using `json_schema` is recommended for models that support it. Note that the

1330 model will not generate JSON without a system or user message instructing it

1331 to do so.

1775 1332

1776 The ranking options for the file search. If not specified, the file search tool will use the `auto` ranker and a score_threshold of 0.1333 - `JsonValue; type "json_object"constant`

1777 1334

1778 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.1335 The type of response format being defined. Always `json_object`.

1779 1336

1780 - `double scoreThreshold`1337 - `JSON_OBJECT("json_object")`

~~1781~~

1782 The score threshold for the file search. All values must be a floating point number between 0 and 1.

~~1783~~

1784 - `Optional<Ranker> ranker`

~~1785~~

1786 The ranker to use for the file search. If not specified will use the `auto` ranker.

~~1787~~

1788 - `AUTO("auto")`

~~1789~~

1790 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~1791~~

1792 - `class FunctionTool:`

~~1793~~

1794 - `FunctionDefinition function`

~~1795~~

1796 - `String name`

~~1797~~

1798 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.

~~1799~~

1800 - `Optional<String> description`

~~1801~~

1802 A description of what the function does, used by the model to choose when and how to call the function.

~~1803~~

1804 - `Optional<FunctionParameters> parameters`

~~1805~~

1806 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~1807~~

1808 Omitting `parameters` defines a function with an empty parameter list.

~~1809~~

1810 - `Optional<Boolean> strict`

~~1811~~

1812 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](https://platform.openai.com/docs/guides/function-calling).

~~1813~~

1814 - `JsonValue; type "function"constant`

~~1815~~

1816 The type of tool being defined: `function`

~~1817~~

1818 - `FUNCTION("function")`

~~1819~~

1820 - `Optional<AssistantResponseFormatOption> responseFormat`

~~1821~~

1822 Specifies the format that the model must output. Compatible with [GPT-4o](https://platform.openai.com/docs/models#gpt-4o), [GPT-4 Turbo](https://platform.openai.com/docs/models#gpt-4-turbo-and-gpt-4), and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.

~~1823~~

1824 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](https://platform.openai.com/docs/guides/structured-outputs).

~~1825~~

1826 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

~~1827~~

1828 **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.

~~1829~~

1830 - `JsonValue;`

~~1831~~

1832 - `AUTO("auto")`

~~1833~~

1834 - `class ResponseFormatText:`

~~1835~~

1836 Default response format. Used to generate text responses.

~~1837~~

1838 - `JsonValue; type "text"constant`

~~1839~~

1840 The type of response format being defined. Always `text`.

~~1841~~

1842 - `TEXT("text")`

~~1843~~

1844 - `class ResponseFormatJsonObject:`

~~1845~~

1846 JSON object response format. An older method of generating JSON responses.

1847 Using `json_schema` is recommended for models that support it. Note that the

1848 model will not generate JSON without a system or user message instructing it

1849 to do so.

~~1850~~

1851 - `JsonValue; type "json_object"constant`

~~1852~~

1853 The type of response format being defined. Always `json_object`.

~~1854~~

1855 - `JSON_OBJECT("json_object")`

1856 1338

1857 - `class ResponseFormatJsonSchema:`1339 - `class ResponseFormatJsonSchema:`

1858 1340

2225 1707

2226 The Unix timestamp (in seconds) for when the run was started.1708 The Unix timestamp (in seconds) for when the run was started.

2227 1709

2228 - `RunStatus status`1710 - `JsonValue status`

~~2229~~

2230 The status of the run, which can be either `queued`, `in_progress`, `requires_action`, `cancelling`, `cancelled`, `failed`, `completed`, `incomplete`, or `expired`.

~~2231~~

2232 - `QUEUED("queued")`

~~2233~~

2234 - `IN_PROGRESS("in_progress")`

~~2235~~

2236 - `REQUIRES_ACTION("requires_action")`

~~2237~~

2238 - `CANCELLING("cancelling")`

~~2239~~

2240 - `CANCELLED("cancelled")`

~~2241~~

2242 - `FAILED("failed")`

~~2243~~

2244 - `COMPLETED("completed")`

~~2245~~

2246 - `INCOMPLETE("incomplete")`

~~2247~~

2248 - `EXPIRED("expired")`

2249 1711

2250 - `String threadId`1712 - `String threadId`

2251 1713

2287 1749

2288 The name of the function to call.1750 The name of the function to call.

2289 1751

2290 - `List<AssistantTool> tools`1752 - `List<JsonValue> tools`

2291 1753

2292 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.1754 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

2293 1755

2294 - `class CodeInterpreterTool:`

~~2295~~

2296 - `JsonValue; type "code_interpreter"constant`

~~2297~~

2298 The type of tool being defined: `code_interpreter`

~~2299~~

2300 - `CODE_INTERPRETER("code_interpreter")`

~~2301~~

2302 - `class FileSearchTool:`

~~2303~~

2304 - `JsonValue; type "file_search"constant`

~~2305~~

2306 The type of tool being defined: `file_search`

~~2307~~

2308 - `FILE_SEARCH("file_search")`

~~2309~~

2310 - `Optional<FileSearch> fileSearch`

~~2311~~

2312 Overrides for the file search tool.

~~2313~~

2314 - `Optional<Long> maxNumResults`

~~2315~~

2316 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.

~~2317~~

2318 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~2319~~

2320 - `Optional<RankingOptions> rankingOptions`

~~2321~~

2322 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.

~~2323~~

2324 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~2325~~

2326 - `double scoreThreshold`

~~2327~~

2328 The score threshold for the file search. All values must be a floating point number between 0 and 1.

~~2329~~

2330 - `Optional<Ranker> ranker`

~~2331~~

2332 The ranker to use for the file search. If not specified will use the `auto` ranker.

~~2333~~

2334 - `AUTO("auto")`

~~2335~~

2336 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~2337~~

2338 - `class FunctionTool:`

~~2339~~

2340 - `FunctionDefinition function`

~~2341~~

2342 - `String name`

~~2343~~

2344 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.

~~2345~~

2346 - `Optional<String> description`

~~2347~~

2348 A description of what the function does, used by the model to choose when and how to call the function.

~~2349~~

2350 - `Optional<FunctionParameters> parameters`

~~2351~~

2352 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~2353~~

2354 Omitting `parameters` defines a function with an empty parameter list.

~~2355~~

2356 - `Optional<Boolean> strict`

~~2357~~

2358 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](https://platform.openai.com/docs/guides/function-calling).

~~2359~~

2360 - `JsonValue; type "function"constant`

~~2361~~

2362 The type of tool being defined: `function`

~~2363~~

2364 - `FUNCTION("function")`

~~2365~~

2366 - `Optional<TruncationStrategy> truncationStrategy`1756 - `Optional<TruncationStrategy> truncationStrategy`

2367 1757

2368 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.1758 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

2604 1994

2605 Details of the tool call.1995 Details of the tool call.

2606 1996

2607 - `List<ToolCall> toolCalls`1997 - `List<JsonValue> toolCalls`

2608 1998

2609 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`.1999 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`.

2610 2000

2611 - `class CodeInterpreterToolCall:`2001 - `JsonValue; type "tool_calls"constant`

~~2612~~

2613 Details of the Code Interpreter tool call the run step was involved in.

2614 2002

2615 - `String id`2003 Always `tool_calls`.

2616 2004

2617 The ID of the tool call.2005 - `TOOL_CALLS("tool_calls")`

2618 2006

2619 - `CodeInterpreter codeInterpreter`2007 - `String threadId`

2620 2008

2621 The Code Interpreter tool call definition.2009 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.

2622 2010

2623 - `String input`2011 - `Type type`

2624 2012

2625 The input to the Code Interpreter tool call.2013 The type of run step, which can be either `message_creation` or `tool_calls`.

2626 2014

2627 - `List<Output> outputs`2015 - `MESSAGE_CREATION("message_creation")`

2628 2016

2629 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.2017 - `TOOL_CALLS("tool_calls")`

2630 2018

2631 - `class LogsOutput:`2019 - `Optional<Usage> usage`

2632 2020

2633 Text output from the Code Interpreter tool call as part of a run step.2021 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.

2634 2022

2635 - `String logs`2023 - `long completionTokens`

2636 2024

2637 The text output from the Code Interpreter tool call.2025 Number of completion tokens used over the course of the run step.

2638 2026

2639 - `JsonValue; type "logs"constant`2027 - `long promptTokens`

2640 2028

2641 Always `logs`.2029 Number of prompt tokens used over the course of the run step.

2642 2030

2643 - `LOGS("logs")`2031 - `long totalTokens`

2644 2032

2645 - `class ImageOutput:`2033 Total number of tokens used (prompt + completion).

2646 2034

2647 - `Image image`2035 - `JsonValue; event "thread.run.step.created"constant`

2648 2036

2649 - `String fileId`2037 - `THREAD_RUN_STEP_CREATED("thread.run.step.created")`

2650 2038

2651 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.2039 - `ThreadRunStepInProgress`

2652 2040

2653 - `JsonValue; type "image"constant`2041 - `RunStep data`

2654 2042

2655 Always `image`.2043 Represents a step in execution of a run.

2656 2044

2657 - `IMAGE("image")`2045 - `JsonValue; event "thread.run.step.in_progress"constant`

2658 2046

2659 - `JsonValue; type "code_interpreter"constant`2047 - `THREAD_RUN_STEP_IN_PROGRESS("thread.run.step.in_progress")`

2660 2048

2661 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.2049 - `ThreadRunStepDelta`

2662 2050

2663 - `CODE_INTERPRETER("code_interpreter")`2051 - `RunStepDeltaEvent data`

2664 2052

2665 - `class FileSearchToolCall:`2053 Represents a run step delta i.e. any changed fields on a run step during streaming.

2666 2054

2667 - `String id`2055 - `String id`

2668 2056

2669 The ID of the tool call object.2057 The identifier of the run step, which can be referenced in API endpoints.

2670 2058

2671 - `FileSearch fileSearch`2059 - `JsonValue delta`

2672 2060

2673 For now, this is always going to be an empty object.2061 - `JsonValue; object_ "thread.run.step.delta"constant`

2674 2062

2675 - `Optional<RankingOptions> rankingOptions`2063 The object type, which is always `thread.run.step.delta`.

2676 2064

2677 The ranking options for the file search.2065 - `THREAD_RUN_STEP_DELTA("thread.run.step.delta")`

2678 2066

2679 - `Ranker ranker`2067 - `JsonValue; event "thread.run.step.delta"constant`

2680 2068

2681 The ranker to use for the file search. If not specified will use the `auto` ranker.2069 - `THREAD_RUN_STEP_DELTA("thread.run.step.delta")`

2682 2070

2683 - `AUTO("auto")`2071 - `ThreadRunStepCompleted`

2684 2072

2685 - `DEFAULT_2024_08_21("default_2024_08_21")`2073 - `RunStep data`

2686 2074

2687 - `double scoreThreshold`2075 Represents a step in execution of a run.

2688 2076

2689 The score threshold for the file search. All values must be a floating point number between 0 and 1.2077 - `JsonValue; event "thread.run.step.completed"constant`

2690 2078

2691 - `Optional<List<Result>> results`2079 - `THREAD_RUN_STEP_COMPLETED("thread.run.step.completed")`

2692 2080

2693 The results of the file search.2081 - `ThreadRunStepFailed`

2694 2082

2695 - `String fileId`2083 - `RunStep data`

2696 2084

2697 The ID of the file that result was found in.2085 Represents a step in execution of a run.

2698 2086

2699 - `String fileName`2087 - `JsonValue; event "thread.run.step.failed"constant`

2700 2088

2701 The name of the file that result was found in.2089 - `THREAD_RUN_STEP_FAILED("thread.run.step.failed")`

2702 2090

2703 - `double score`2091 - `ThreadRunStepCancelled`

2704 2092

2705 The score of the result. All values must be a floating point number between 0 and 1.2093 - `RunStep data`

2706 2094

2707 - `Optional<List<Content>> content`2095 Represents a step in execution of a run.

2708 2096

2709 The content of the result that was found. The content is only included if requested via the include query parameter.2097 - `JsonValue; event "thread.run.step.cancelled"constant`

2710 2098

2711 - `Optional<String> text`2099 - `THREAD_RUN_STEP_CANCELLED("thread.run.step.cancelled")`

2712 2100

2713 The text content of the file.2101 - `ThreadRunStepExpired`

2714 2102

2715 - `Optional<Type> type`2103 - `RunStep data`

2716 2104

2717 The type of the content.2105 Represents a step in execution of a run.

2718 2106

2719 - `TEXT("text")`2107 - `JsonValue; event "thread.run.step.expired"constant`

2720 2108

2721 - `JsonValue; type "file_search"constant`2109 - `THREAD_RUN_STEP_EXPIRED("thread.run.step.expired")`

2722 2110

2723 The type of tool call. This is always going to be `file_search` for this type of tool call.2111 - `ThreadMessageCreated`

2724 2112

2725 - `FILE_SEARCH("file_search")`2113 - `Message data`

2726 2114

2727 - `class FunctionToolCall:`2115 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2728 2116

2729 - `String id`2117 - `String id`

2730 2118

2731 The ID of the tool call object.2119 The identifier, which can be referenced in API endpoints.

2732 2120

2733 - `Function function`2121 - `Optional<String> assistantId`

2734 2122

2735 The definition of the function that was called.2123 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.

2736 2124

2737 - `String arguments`2125 - `Optional<List<Attachment>> attachments`

2738 2126

2739 The arguments passed to the function.2127 A list of files attached to the message, and the tools they were added to.

2740 2128

2741 - `String name`2129 - `Optional<String> fileId`

2742 2130

2743 The name of the function.2131 The ID of the file to attach to the message.

2744 2132

2745 - `Optional<String> output`2133 - `Optional<List<Tool>> tools`

2746 2134

2747 The output of the function. This will be `null` if the outputs have not been [submitted](https://platform.openai.com/docs/api-reference/runs/submitToolOutputs) yet.2135 The tools to add this file to.

2748 2136

2749 - `JsonValue; type "function"constant`2137 - `class CodeInterpreterTool:`

2750 2138

2751 The type of tool call. This is always going to be `function` for this type of tool call.2139 - `JsonValue; type "code_interpreter"constant`

2752 2140

2753 - `FUNCTION("function")`2141 The type of tool being defined: `code_interpreter`

2754 2142

2755 - `JsonValue; type "tool_calls"constant`2143 - `CODE_INTERPRETER("code_interpreter")`

2756 2144

2757 Always `tool_calls`.2145 - `JsonValue;`

2758 2146

2759 - `TOOL_CALLS("tool_calls")`2147 - `JsonValue; type "file_search"constant`

2760 2148

2761 - `String threadId`2149 The type of tool being defined: `file_search`

2762 2150

2763 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.2151 - `FILE_SEARCH("file_search")`

~~2764~~

2765 - `Type type`

~~2766~~

2767 The type of run step, which can be either `message_creation` or `tool_calls`.

~~2768~~

2769 - `MESSAGE_CREATION("message_creation")`

~~2770~~

2771 - `TOOL_CALLS("tool_calls")`

~~2772~~

2773 - `Optional<Usage> usage`

~~2774~~

2775 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.

~~2776~~

2777 - `long completionTokens`

~~2778~~

2779 Number of completion tokens used over the course of the run step.

~~2780~~

2781 - `long promptTokens`

~~2782~~

2783 Number of prompt tokens used over the course of the run step.

~~2784~~

2785 - `long totalTokens`

~~2786~~

2787 Total number of tokens used (prompt + completion).

~~2788~~

2789 - `JsonValue; event "thread.run.step.created"constant`

~~2790~~

2791 - `THREAD_RUN_STEP_CREATED("thread.run.step.created")`

~~2792~~

2793 - `ThreadRunStepInProgress`

~~2794~~

2795 - `RunStep data`

~~2796~~

2797 Represents a step in execution of a run.

~~2798~~

2799 - `JsonValue; event "thread.run.step.in_progress"constant`

~~2800~~

2801 - `THREAD_RUN_STEP_IN_PROGRESS("thread.run.step.in_progress")`

~~2802~~

2803 - `ThreadRunStepDelta`

~~2804~~

2805 - `RunStepDeltaEvent data`

~~2806~~

2807 Represents a run step delta i.e. any changed fields on a run step during streaming.

~~2808~~

2809 - `String id`

~~2810~~

2811 The identifier of the run step, which can be referenced in API endpoints.

~~2812~~

2813 - `RunStepDelta delta`

~~2814~~

2815 The delta containing the fields that have changed on the run step.

~~2816~~

2817 - `Optional<StepDetails> stepDetails`

~~2818~~

2819 The details of the run step.

~~2820~~

2821 - `class RunStepDeltaMessageDelta:`

~~2822~~

2823 Details of the message creation by the run step.

~~2824~~

2825 - `JsonValue; type "message_creation"constant`

~~2826~~

2827 Always `message_creation`.

~~2828~~

2829 - `MESSAGE_CREATION("message_creation")`

~~2830~~

2831 - `Optional<MessageCreation> messageCreation`

~~2832~~

2833 - `Optional<String> messageId`

~~2834~~

2835 The ID of the message that was created by this run step.

~~2836~~

2837 - `class ToolCallDeltaObject:`

~~2838~~

2839 Details of the tool call.

~~2840~~

2841 - `JsonValue; type "tool_calls"constant`

~~2842~~

2843 Always `tool_calls`.

~~2844~~

2845 - `TOOL_CALLS("tool_calls")`

~~2846~~

2847 - `Optional<List<ToolCallDelta>> toolCalls`

~~2848~~

2849 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`.

~~2850~~

2851 - `class CodeInterpreterToolCallDelta:`

~~2852~~

2853 Details of the Code Interpreter tool call the run step was involved in.

~~2854~~

2855 - `long index`

~~2856~~

2857 The index of the tool call in the tool calls array.

~~2858~~

2859 - `JsonValue; type "code_interpreter"constant`

~~2860~~

2861 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.

~~2862~~

2863 - `CODE_INTERPRETER("code_interpreter")`

~~2864~~

2865 - `Optional<String> id`

~~2866~~

2867 The ID of the tool call.

~~2868~~

2869 - `Optional<CodeInterpreter> codeInterpreter`

~~2870~~

2871 The Code Interpreter tool call definition.

~~2872~~

2873 - `Optional<String> input`

~~2874~~

2875 The input to the Code Interpreter tool call.

~~2876~~

2877 - `Optional<List<Output>> outputs`

~~2878~~

2879 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.

~~2880~~

2881 - `class CodeInterpreterLogs:`

~~2882~~

2883 Text output from the Code Interpreter tool call as part of a run step.

~~2884~~

2885 - `long index`

~~2886~~

2887 The index of the output in the outputs array.

~~2888~~

2889 - `JsonValue; type "logs"constant`

~~2890~~

2891 Always `logs`.

~~2892~~

2893 - `LOGS("logs")`

~~2894~~

2895 - `Optional<String> logs`

~~2896~~

2897 The text output from the Code Interpreter tool call.

~~2898~~

2899 - `class CodeInterpreterOutputImage:`

~~2900~~

2901 - `long index`

~~2902~~

2903 The index of the output in the outputs array.

~~2904~~

2905 - `JsonValue; type "image"constant`

~~2906~~

2907 Always `image`.

~~2908~~

2909 - `IMAGE("image")`

~~2910~~

2911 - `Optional<Image> image`

~~2912~~

2913 - `Optional<String> fileId`

~~2914~~

2915 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.

~~2916~~

2917 - `class FileSearchToolCallDelta:`

~~2918~~

2919 - `JsonValue fileSearch`

~~2920~~

2921 For now, this is always going to be an empty object.

~~2922~~

2923 - `long index`

~~2924~~

2925 The index of the tool call in the tool calls array.

~~2926~~

2927 - `JsonValue; type "file_search"constant`

~~2928~~

2929 The type of tool call. This is always going to be `file_search` for this type of tool call.

~~2930~~

2931 - `FILE_SEARCH("file_search")`

~~2932~~

2933 - `Optional<String> id`

~~2934~~

2935 The ID of the tool call object.

~~2936~~

2937 - `class FunctionToolCallDelta:`

~~2938~~

2939 - `long index`

~~2940~~

2941 The index of the tool call in the tool calls array.

~~2942~~

2943 - `JsonValue; type "function"constant`

~~2944~~

2945 The type of tool call. This is always going to be `function` for this type of tool call.

~~2946~~

2947 - `FUNCTION("function")`

~~2948~~

2949 - `Optional<String> id`

~~2950~~

2951 The ID of the tool call object.

~~2952~~

2953 - `Optional<Function> function`

~~2954~~

2955 The definition of the function that was called.

~~2956~~

2957 - `Optional<String> arguments`

~~2958~~

2959 The arguments passed to the function.

~~2960~~

2961 - `Optional<String> name`

~~2962~~

2963 The name of the function.

~~2964~~

2965 - `Optional<String> output`

~~2966~~

2967 The output of the function. This will be `null` if the outputs have not been [submitted](https://platform.openai.com/docs/api-reference/runs/submitToolOutputs) yet.

~~2968~~

2969 - `JsonValue; object_ "thread.run.step.delta"constant`

~~2970~~

2971 The object type, which is always `thread.run.step.delta`.

~~2972~~

2973 - `THREAD_RUN_STEP_DELTA("thread.run.step.delta")`

~~2974~~

2975 - `JsonValue; event "thread.run.step.delta"constant`

~~2976~~

2977 - `THREAD_RUN_STEP_DELTA("thread.run.step.delta")`

~~2978~~

2979 - `ThreadRunStepCompleted`

~~2980~~

2981 - `RunStep data`

~~2982~~

2983 Represents a step in execution of a run.

~~2984~~

2985 - `JsonValue; event "thread.run.step.completed"constant`

~~2986~~

2987 - `THREAD_RUN_STEP_COMPLETED("thread.run.step.completed")`

~~2988~~

2989 - `ThreadRunStepFailed`

~~2990~~

2991 - `RunStep data`

~~2992~~

2993 Represents a step in execution of a run.

~~2994~~

2995 - `JsonValue; event "thread.run.step.failed"constant`

~~2996~~

2997 - `THREAD_RUN_STEP_FAILED("thread.run.step.failed")`

~~2998~~

2999 - `ThreadRunStepCancelled`

~~3000~~

3001 - `RunStep data`

~~3002~~

3003 Represents a step in execution of a run.

~~3004~~

3005 - `JsonValue; event "thread.run.step.cancelled"constant`

~~3006~~

3007 - `THREAD_RUN_STEP_CANCELLED("thread.run.step.cancelled")`

~~3008~~

3009 - `ThreadRunStepExpired`

~~3010~~

3011 - `RunStep data`

~~3012~~

3013 Represents a step in execution of a run.

~~3014~~

3015 - `JsonValue; event "thread.run.step.expired"constant`

~~3016~~

3017 - `THREAD_RUN_STEP_EXPIRED("thread.run.step.expired")`

~~3018~~

3019 - `ThreadMessageCreated`

~~3020~~

3021 - `Message data`

~~3022~~

3023 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

~~3024~~

3025 - `String id`

~~3026~~

3027 The identifier, which can be referenced in API endpoints.

~~3028~~

3029 - `Optional<String> assistantId`

~~3030~~

3031 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.

~~3032~~

3033 - `Optional<List<Attachment>> attachments`

~~3034~~

3035 A list of files attached to the message, and the tools they were added to.

~~3036~~

3037 - `Optional<String> fileId`

~~3038~~

3039 The ID of the file to attach to the message.

~~3040~~

3041 - `Optional<List<Tool>> tools`

~~3042~~

3043 The tools to add this file to.

~~3044~~

3045 - `class CodeInterpreterTool:`

~~3046~~

3047 - `JsonValue;`

~~3048~~

3049 - `JsonValue; type "file_search"constant`

~~3050~~

3051 The type of tool being defined: `file_search`

~~3052~~

3053 - `FILE_SEARCH("file_search")`

~~3054~~

3055 - `Optional<Long> completedAt`

~~3056~~

3057 The Unix timestamp (in seconds) for when the message was completed.

~~3058~~

3059 - `List<MessageContent> content`

~~3060~~

3061 The content of the message in array of text and/or images.

~~3062~~

3063 - `class ImageFileContentBlock:`

~~3064~~

3065 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

~~3066~~

3067 - `ImageFile imageFile`

~~3068~~

3069 - `String fileId`

~~3070~~

3071 The [File](https://platform.openai.com/docs/api-reference/files) ID of the image in the message content. Set `purpose="vision"` when uploading the File if you need to later display the file content.

~~3072~~

3073 - `Optional<Detail> detail`

~~3074~~

3075 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`.

~~3076~~

3077 - `AUTO("auto")`

~~3078~~

3079 - `LOW("low")`

~~3080~~

3081 - `HIGH("high")`

~~3082~~

3083 - `JsonValue; type "image_file"constant`

~~3084~~

3085 Always `image_file`.

~~3086~~

3087 - `IMAGE_FILE("image_file")`

~~3088~~

3089 - `class ImageUrlContentBlock:`

~~3090~~

3091 References an image URL in the content of a message.

~~3092~~

3093 - `ImageUrl imageUrl`

~~3094~~

3095 - `String url`

~~3096~~

3097 The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

~~3098~~

3099 - `Optional<Detail> detail`

~~3100~~

3101 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`. Default value is `auto`

~~3102~~

3103 - `AUTO("auto")`

~~3104~~

3105 - `LOW("low")`

~~3106~~

3107 - `HIGH("high")`

~~3108~~

3109 - `JsonValue; type "image_url"constant`

~~3110~~

3111 The type of the content part.

~~3112~~

3113 - `IMAGE_URL("image_url")`

~~3114~~

3115 - `class TextContentBlock:`

~~3116~~

3117 The text content that is part of a message.

~~3118~~

3119 - `Text text`

~~3120~~

3121 - `List<Annotation> annotations`

~~3122~~

3123 - `class FileCitationAnnotation:`

~~3124~~

3125 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.

~~3126~~

3127 - `long endIndex`

~~3128~~

3129 - `FileCitation fileCitation`

~~3130~~

3131 - `String fileId`

~~3132~~

3133 The ID of the specific File the citation is from.

~~3134~~

3135 - `long startIndex`

~~3136~~

3137 - `String text`

~~3138~~

3139 The text in the message content that needs to be replaced.

~~3140~~

3141 - `JsonValue; type "file_citation"constant`

~~3142~~

3143 Always `file_citation`.

~~3144~~

3145 - `FILE_CITATION("file_citation")`

~~3146~~

3147 - `class FilePathAnnotation:`

~~3148~~

3149 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.

~~3150~~

3151 - `long endIndex`

~~3152~~

3153 - `FilePath filePath`

~~3154~~

3155 - `String fileId`

~~3156~~

3157 The ID of the file that was generated.

~~3158~~

3159 - `long startIndex`

~~3160~~

3161 - `String text`

~~3162~~

3163 The text in the message content that needs to be replaced.

~~3164~~

3165 - `JsonValue; type "file_path"constant`

~~3166~~

3167 Always `file_path`.

~~3168~~

3169 - `FILE_PATH("file_path")`

~~3170~~

3171 - `String value`

~~3172~~

3173 The data that makes up the text.

~~3174~~

3175 - `JsonValue; type "text"constant`

~~3176~~

3177 Always `text`.

~~3178~~

3179 - `TEXT("text")`

~~3180~~

3181 - `class RefusalContentBlock:`

~~3182~~

3183 The refusal content generated by the assistant.

~~3184~~

3185 - `String refusal`

~~3186~~

3187 - `JsonValue; type "refusal"constant`

~~3188~~

3189 Always `refusal`.

~~3190~~

3191 - `REFUSAL("refusal")`

~~3192~~

3193 - `long createdAt`

~~3194~~

3195 The Unix timestamp (in seconds) for when the message was created.

~~3196~~

3197 - `Optional<Long> incompleteAt`

~~3198~~

3199 The Unix timestamp (in seconds) for when the message was marked as incomplete.

~~3200~~

3201 - `Optional<IncompleteDetails> incompleteDetails`

~~3202~~

3203 On an incomplete message, details about why the message is incomplete.

~~3204~~

3205 - `Reason reason`

~~3206~~

3207 The reason the message is incomplete.

~~3208~~

3209 - `CONTENT_FILTER("content_filter")`

~~3210~~

3211 - `MAX_TOKENS("max_tokens")`

~~3212~~

3213 - `RUN_CANCELLED("run_cancelled")`

~~3214~~

3215 - `RUN_EXPIRED("run_expired")`

~~3216~~

3217 - `RUN_FAILED("run_failed")`

~~3218~~

3219 - `Optional<Metadata> metadata`

~~3220~~

3221 Set of 16 key-value pairs that can be attached to an object. This can be

3222 useful for storing additional information about the object in a structured

3223 format, and querying for objects via API or the dashboard.

~~3224~~

3225 Keys are strings with a maximum length of 64 characters. Values are strings

3226 with a maximum length of 512 characters.

~~3227~~

3228 - `JsonValue; object_ "thread.message"constant`

~~3229~~

3230 The object type, which is always `thread.message`.

~~3231~~

3232 - `THREAD_MESSAGE("thread.message")`

~~3233~~

3234 - `Role role`

~~3235~~

3236 The entity that produced the message. One of `user` or `assistant`.

~~3237~~

3238 - `USER("user")`

~~3239~~

3240 - `ASSISTANT("assistant")`

~~3241~~

3242 - `Optional<String> runId`

~~3243~~

3244 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) associated with the creation of this message. Value is `null` when messages are created manually using the create message or create thread endpoints.

~~3245~~

3246 - `Status status`

~~3247~~

3248 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

~~3249~~

3250 - `IN_PROGRESS("in_progress")`

~~3251~~

3252 - `INCOMPLETE("incomplete")`

~~3253~~

3254 - `COMPLETED("completed")`

~~3255~~

3256 - `String threadId`

~~3257~~

3258 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.

~~3259~~

3260 - `JsonValue; event "thread.message.created"constant`

~~3261~~

3262 - `THREAD_MESSAGE_CREATED("thread.message.created")`

~~3263~~

3264 - `ThreadMessageInProgress`

~~3265~~

3266 - `Message data`

~~3267~~

3268 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

~~3269~~

3270 - `JsonValue; event "thread.message.in_progress"constant`

~~3271~~

3272 - `THREAD_MESSAGE_IN_PROGRESS("thread.message.in_progress")`

~~3273~~

3274 - `ThreadMessageDelta`

~~3275~~

3276 - `MessageDeltaEvent data`

~~3277~~

3278 Represents a message delta i.e. any changed fields on a message during streaming.

~~3279~~

3280 - `String id`

~~3281~~

3282 The identifier of the message, which can be referenced in API endpoints.

~~3283~~

3284 - `MessageDelta delta`

~~3285~~

3286 The delta containing the fields that have changed on the Message.

~~3287~~

3288 - `Optional<List<MessageContentDelta>> content`

~~3289~~

3290 The content of the message in array of text and/or images.

~~3291~~

3292 - `class ImageFileDeltaBlock:`

~~3293~~

3294 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

~~3295~~

3296 - `long index`

~~3297~~

3298 The index of the content part in the message.

~~3299~~

3300 - `JsonValue; type "image_file"constant`

~~3301~~

3302 Always `image_file`.

~~3303~~

3304 - `IMAGE_FILE("image_file")`

~~3305~~

3306 - `Optional<ImageFileDelta> imageFile`

~~3307~~

3308 - `Optional<Detail> detail`

~~3309~~

3310 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`.

~~3311~~

3312 - `AUTO("auto")`

~~3313~~

3314 - `LOW("low")`

~~3315~~

3316 - `HIGH("high")`

~~3317~~

3318 - `Optional<String> fileId`

~~3319~~

3320 The [File](https://platform.openai.com/docs/api-reference/files) ID of the image in the message content. Set `purpose="vision"` when uploading the File if you need to later display the file content.

~~3321~~

3322 - `class TextDeltaBlock:`

~~3323~~

3324 The text content that is part of a message.

~~3325~~

3326 - `long index`

~~3327~~

3328 The index of the content part in the message.

~~3329~~

3330 - `JsonValue; type "text"constant`

~~3331~~

3332 Always `text`.

~~3333~~

3334 - `TEXT("text")`

~~3335~~

3336 - `Optional<TextDelta> text`

~~3337~~

3338 - `Optional<List<AnnotationDelta>> annotations`

~~3339~~

3340 - `class FileCitationDeltaAnnotation:`

~~3341~~

3342 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.

~~3343~~

3344 - `long index`

~~3345~~

3346 The index of the annotation in the text content part.

~~3347~~

3348 - `JsonValue; type "file_citation"constant`

~~3349~~

3350 Always `file_citation`.

~~3351~~

3352 - `FILE_CITATION("file_citation")`

~~3353~~

3354 - `Optional<Long> endIndex`

~~3355~~

3356 - `Optional<FileCitation> fileCitation`

~~3357~~

3358 - `Optional<String> fileId`

~~3359~~

3360 The ID of the specific File the citation is from.

~~3361~~

3362 - `Optional<String> quote`

~~3363~~

3364 The specific quote in the file.

~~3365~~

3366 - `Optional<Long> startIndex`

~~3367~~

3368 - `Optional<String> text`

~~3369~~

3370 The text in the message content that needs to be replaced.

~~3371~~

3372 - `class FilePathDeltaAnnotation:`

~~3373~~

3374 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.

~~3375~~

3376 - `long index`

~~3377~~

3378 The index of the annotation in the text content part.

~~3379~~

3380 - `JsonValue; type "file_path"constant`

~~3381~~

3382 Always `file_path`.

~~3383~~

3384 - `FILE_PATH("file_path")`

~~3385~~

3386 - `Optional<Long> endIndex`

~~3387~~

3388 - `Optional<FilePath> filePath`

~~3389~~

3390 - `Optional<String> fileId`

~~3391~~

3392 The ID of the file that was generated.

~~3393~~

3394 - `Optional<Long> startIndex`

~~3395~~

3396 - `Optional<String> text`

~~3397~~

3398 The text in the message content that needs to be replaced.

~~3399~~

3400 - `Optional<String> value`

~~3401~~

3402 The data that makes up the text.

~~3403~~

3404 - `class RefusalDeltaBlock:`

~~3405~~

3406 The refusal content that is part of a message.

~~3407~~

3408 - `long index`

~~3409~~

3410 The index of the refusal part in the message.

~~3411~~

3412 - `JsonValue; type "refusal"constant`

~~3413~~

3414 Always `refusal`.

~~3415~~

3416 - `REFUSAL("refusal")`

~~3417~~

3418 - `Optional<String> refusal`

~~3419~~

3420 - `class ImageUrlDeltaBlock:`

~~3421~~

3422 References an image URL in the content of a message.

~~3423~~

3424 - `long index`

~~3425~~

3426 The index of the content part in the message.

~~3427~~

3428 - `JsonValue; type "image_url"constant`

~~3429~~

3430 Always `image_url`.

~~3431~~

3432 - `IMAGE_URL("image_url")`

~~3433~~

3434 - `Optional<ImageUrlDelta> imageUrl`

~~3435~~

3436 - `Optional<Detail> detail`

~~3437~~

3438 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`.

~~3439~~

3440 - `AUTO("auto")`

~~3441~~

3442 - `LOW("low")`

~~3443~~

3444 - `HIGH("high")`

~~3445~~

3446 - `Optional<String> url`

~~3447~~

3448 The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

~~3449~~

3450 - `Optional<Role> role`

~~3451~~

3452 The entity that produced the message. One of `user` or `assistant`.

~~3453~~

3454 - `USER("user")`

~~3455~~

3456 - `ASSISTANT("assistant")`

~~3457~~

3458 - `JsonValue; object_ "thread.message.delta"constant`

~~3459~~

3460 The object type, which is always `thread.message.delta`.

~~3461~~

3462 - `THREAD_MESSAGE_DELTA("thread.message.delta")`

~~3463~~

3464 - `JsonValue; event "thread.message.delta"constant`

~~3465~~

3466 - `THREAD_MESSAGE_DELTA("thread.message.delta")`

~~3467~~

3468 - `ThreadMessageCompleted`

~~3469~~

3470 - `Message data`

~~3471~~

3472 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

~~3473~~

3474 - `JsonValue; event "thread.message.completed"constant`

~~3475~~

3476 - `THREAD_MESSAGE_COMPLETED("thread.message.completed")`

~~3477~~

3478 - `ThreadMessageIncomplete`

~~3479~~

3480 - `Message data`

~~3481~~

3482 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

~~3483~~

3484 - `JsonValue; event "thread.message.incomplete"constant`

~~3485~~

3486 - `THREAD_MESSAGE_INCOMPLETE("thread.message.incomplete")`

~~3487~~

3488 - `ErrorEvent`

~~3489~~

3490 - `ErrorObject data`

~~3491~~

3492 - `Optional<String> code`

~~3493~~

3494 - `String message`

~~3495~~

3496 - `Optional<String> param`

~~3497~~

3498 - `String type`

~~3499~~

3500 - `JsonValue; event "error"constant`

~~3501~~

3502 - `ERROR("error")`

~~3503~~

3504### Assistant Tool

~~3505~~

3506- `class AssistantTool: A class that can be one of several variants.union`

~~3507~~

3508 - `class CodeInterpreterTool:`

~~3509~~

3510 - `JsonValue; type "code_interpreter"constant`

~~3511~~

3512 The type of tool being defined: `code_interpreter`

~~3513~~

3514 - `CODE_INTERPRETER("code_interpreter")`

~~3515~~

3516 - `class FileSearchTool:`

~~3517~~

3518 - `JsonValue; type "file_search"constant`

~~3519~~

3520 The type of tool being defined: `file_search`

~~3521~~

3522 - `FILE_SEARCH("file_search")`

~~3523~~

3524 - `Optional<FileSearch> fileSearch`

~~3525~~

3526 Overrides for the file search tool.

~~3527~~

3528 - `Optional<Long> maxNumResults`

~~3529~~

3530 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.

~~3531~~

3532 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~3533~~

3534 - `Optional<RankingOptions> rankingOptions`

~~3535~~

3536 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.

~~3537~~

3538 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~3539~~

3540 - `double scoreThreshold`

~~3541~~

3542 The score threshold for the file search. All values must be a floating point number between 0 and 1.

~~3543~~

3544 - `Optional<Ranker> ranker`

~~3545~~

3546 The ranker to use for the file search. If not specified will use the `auto` ranker.

~~3547~~

3548 - `AUTO("auto")`

~~3549~~

3550 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~3551~~

3552 - `class FunctionTool:`

~~3553~~

3554 - `FunctionDefinition function`

~~3555~~

3556 - `String name`

~~3557~~

3558 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.

~~3559~~

3560 - `Optional<String> description`

~~3561~~

3562 A description of what the function does, used by the model to choose when and how to call the function.

~~3563~~

3564 - `Optional<FunctionParameters> parameters`

~~3565~~

3566 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~3567~~

3568 Omitting `parameters` defines a function with an empty parameter list.

~~3569~~

3570 - `Optional<Boolean> strict`

~~3571~~

3572 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](https://platform.openai.com/docs/guides/function-calling).

~~3573~~

3574 - `JsonValue; type "function"constant`

~~3575~~

3576 The type of tool being defined: `function`

~~3577~~

3578 - `FUNCTION("function")`

~~3579~~

3580### Code Interpreter Tool

~~3581~~

3582- `class CodeInterpreterTool:`

~~3583~~

3584 - `JsonValue; type "code_interpreter"constant`

~~3585~~

3586 The type of tool being defined: `code_interpreter`

~~3587~~

3588 - `CODE_INTERPRETER("code_interpreter")`

~~3589~~

3590### File Search Tool

~~3591~~

3592- `class FileSearchTool:`

~~3593~~

3594 - `JsonValue; type "file_search"constant`

~~3595~~

3596 The type of tool being defined: `file_search`

~~3597~~

3598 - `FILE_SEARCH("file_search")`

~~3599~~

3600 - `Optional<FileSearch> fileSearch`

~~3601~~

3602 Overrides for the file search tool.

~~3603~~

3604 - `Optional<Long> maxNumResults`

~~3605~~

3606 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.

~~3607~~

3608 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~3609~~

3610 - `Optional<RankingOptions> rankingOptions`

~~3611~~

3612 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.

~~3613~~

3614 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~3615~~

3616 - `double scoreThreshold`

~~3617~~

3618 The score threshold for the file search. All values must be a floating point number between 0 and 1.

~~3619~~

3620 - `Optional<Ranker> ranker`

~~3621~~

3622 The ranker to use for the file search. If not specified will use the `auto` ranker.

~~3623~~

3624 - `AUTO("auto")`

~~3625~~

3626 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~3627~~

3628### Function Tool

~~3629~~

3630- `class FunctionTool:`

~~3631~~

3632 - `FunctionDefinition function`

~~3633~~

3634 - `String name`

~~3635~~

3636 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.

~~3637~~

3638 - `Optional<String> description`

~~3639~~

3640 A description of what the function does, used by the model to choose when and how to call the function.

~~3641~~

3642 - `Optional<FunctionParameters> parameters`

~~3643~~

3644 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~3645~~

3646 Omitting `parameters` defines a function with an empty parameter list.

~~3647~~

3648 - `Optional<Boolean> strict`

~~3649~~

3650 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](https://platform.openai.com/docs/guides/function-calling).

~~3651~~

3652 - `JsonValue; type "function"constant`

~~3653~~

3654 The type of tool being defined: `function`

~~3655~~

3656 - `FUNCTION("function")`

~~3657~~

3658### Message Stream Event

~~3659~~

3660- `class MessageStreamEvent: A class that can be one of several variants.union`

~~3661~~

3662 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.

~~3663~~

3664 - `ThreadMessageCreated`

~~3665~~

3666 - `Message data`

~~3667~~

3668 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

~~3669~~

3670 - `String id`

~~3671~~

3672 The identifier, which can be referenced in API endpoints.

~~3673~~

3674 - `Optional<String> assistantId`

~~3675~~

3676 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.

~~3677~~

3678 - `Optional<List<Attachment>> attachments`

~~3679~~

3680 A list of files attached to the message, and the tools they were added to.

~~3681~~

3682 - `Optional<String> fileId`

~~3683~~

3684 The ID of the file to attach to the message.

~~3685~~

3686 - `Optional<List<Tool>> tools`

~~3687~~

3688 The tools to add this file to.

~~3689~~

3690 - `class CodeInterpreterTool:`

~~3691~~

3692 - `JsonValue; type "code_interpreter"constant`

~~3693~~

3694 The type of tool being defined: `code_interpreter`

~~3695~~

3696 - `CODE_INTERPRETER("code_interpreter")`

~~3697~~

3698 - `JsonValue;`

~~3699~~

3700 - `JsonValue; type "file_search"constant`

~~3701~~

3702 The type of tool being defined: `file_search`

~~3703~~

3704 - `FILE_SEARCH("file_search")`

~~3705~~

3706 - `Optional<Long> completedAt`

~~3707~~

3708 The Unix timestamp (in seconds) for when the message was completed.

~~3709~~

3710 - `List<MessageContent> content`

~~3711~~

3712 The content of the message in array of text and/or images.

~~3713~~

3714 - `class ImageFileContentBlock:`

~~3715~~

3716 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

~~3717~~

3718 - `ImageFile imageFile`

~~3719~~

3720 - `String fileId`

~~3721~~

3722 The [File](https://platform.openai.com/docs/api-reference/files) ID of the image in the message content. Set `purpose="vision"` when uploading the File if you need to later display the file content.

~~3723~~

3724 - `Optional<Detail> detail`

~~3725~~

3726 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`.

~~3727~~

3728 - `AUTO("auto")`

~~3729~~

3730 - `LOW("low")`

~~3731~~

3732 - `HIGH("high")`

~~3733~~

3734 - `JsonValue; type "image_file"constant`

~~3735~~

3736 Always `image_file`.

~~3737~~

3738 - `IMAGE_FILE("image_file")`

~~3739~~

3740 - `class ImageUrlContentBlock:`

~~3741~~

3742 References an image URL in the content of a message.

~~3743~~

3744 - `ImageUrl imageUrl`

~~3745~~

3746 - `String url`

~~3747~~

3748 The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

~~3749~~

3750 - `Optional<Detail> detail`

~~3751~~

3752 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`. Default value is `auto`

~~3753~~

3754 - `AUTO("auto")`

~~3755~~

3756 - `LOW("low")`

~~3757~~

3758 - `HIGH("high")`

~~3759~~

3760 - `JsonValue; type "image_url"constant`

~~3761~~

3762 The type of the content part.

~~3763~~

3764 - `IMAGE_URL("image_url")`

~~3765~~

3766 - `class TextContentBlock:`

~~3767~~

3768 The text content that is part of a message.

~~3769~~

3770 - `Text text`

~~3771~~

3772 - `List<Annotation> annotations`

~~3773~~

3774 - `class FileCitationAnnotation:`

~~3775~~

3776 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.

~~3777~~

3778 - `long endIndex`

~~3779~~

3780 - `FileCitation fileCitation`

~~3781~~

3782 - `String fileId`

~~3783~~

3784 The ID of the specific File the citation is from.

~~3785~~

3786 - `long startIndex`

~~3787~~

3788 - `String text`

~~3789~~

3790 The text in the message content that needs to be replaced.

~~3791~~

3792 - `JsonValue; type "file_citation"constant`

~~3793~~

3794 Always `file_citation`.

~~3795~~

3796 - `FILE_CITATION("file_citation")`

~~3797~~

3798 - `class FilePathAnnotation:`

~~3799~~

3800 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.

~~3801~~

3802 - `long endIndex`

~~3803~~

3804 - `FilePath filePath`

~~3805~~

3806 - `String fileId`

~~3807~~

3808 The ID of the file that was generated.

~~3809~~

3810 - `long startIndex`

~~3811~~

3812 - `String text`

~~3813~~

3814 The text in the message content that needs to be replaced.

~~3815~~

3816 - `JsonValue; type "file_path"constant`

~~3817~~

3818 Always `file_path`.

~~3819~~

3820 - `FILE_PATH("file_path")`

~~3821~~

3822 - `String value`

~~3823~~

3824 The data that makes up the text.

~~3825~~

3826 - `JsonValue; type "text"constant`

~~3827~~

3828 Always `text`.

~~3829~~

3830 - `TEXT("text")`

~~3831~~

3832 - `class RefusalContentBlock:`

~~3833~~

3834 The refusal content generated by the assistant.

~~3835~~

3836 - `String refusal`

~~3837~~

3838 - `JsonValue; type "refusal"constant`

~~3839~~

3840 Always `refusal`.

~~3841~~

3842 - `REFUSAL("refusal")`

~~3843~~

3844 - `long createdAt`

~~3845~~

3846 The Unix timestamp (in seconds) for when the message was created.

~~3847~~

3848 - `Optional<Long> incompleteAt`

~~3849~~

3850 The Unix timestamp (in seconds) for when the message was marked as incomplete.

~~3851~~

3852 - `Optional<IncompleteDetails> incompleteDetails`

~~3853~~

3854 On an incomplete message, details about why the message is incomplete.

~~3855~~

3856 - `Reason reason`

~~3857~~

3858 The reason the message is incomplete.

~~3859~~

3860 - `CONTENT_FILTER("content_filter")`

~~3861~~

3862 - `MAX_TOKENS("max_tokens")`

~~3863~~

3864 - `RUN_CANCELLED("run_cancelled")`

~~3865~~

3866 - `RUN_EXPIRED("run_expired")`

~~3867~~

3868 - `RUN_FAILED("run_failed")`

~~3869~~

3870 - `Optional<Metadata> metadata`

~~3871~~

3872 Set of 16 key-value pairs that can be attached to an object. This can be

3873 useful for storing additional information about the object in a structured

3874 format, and querying for objects via API or the dashboard.

~~3875~~

3876 Keys are strings with a maximum length of 64 characters. Values are strings

3877 with a maximum length of 512 characters.

~~3878~~

3879 - `JsonValue; object_ "thread.message"constant`

~~3880~~

3881 The object type, which is always `thread.message`.

~~3882~~

3883 - `THREAD_MESSAGE("thread.message")`

~~3884~~

3885 - `Role role`

~~3886~~

3887 The entity that produced the message. One of `user` or `assistant`.

~~3888~~

3889 - `USER("user")`

~~3890~~

3891 - `ASSISTANT("assistant")`

~~3892~~

3893 - `Optional<String> runId`

~~3894~~

3895 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) associated with the creation of this message. Value is `null` when messages are created manually using the create message or create thread endpoints.

~~3896~~

3897 - `Status status`

~~3898~~

3899 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

~~3900~~

3901 - `IN_PROGRESS("in_progress")`

~~3902~~

3903 - `INCOMPLETE("incomplete")`

~~3904~~

3905 - `COMPLETED("completed")`

~~3906~~

3907 - `String threadId`

~~3908~~

3909 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.

~~3910~~

3911 - `JsonValue; event "thread.message.created"constant`

~~3912~~

3913 - `THREAD_MESSAGE_CREATED("thread.message.created")`

~~3914~~

3915 - `ThreadMessageInProgress`

~~3916~~

3917 - `Message data`

~~3918~~

3919 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

~~3920~~

3921 - `JsonValue; event "thread.message.in_progress"constant`

~~3922~~

3923 - `THREAD_MESSAGE_IN_PROGRESS("thread.message.in_progress")`

~~3924~~

3925 - `ThreadMessageDelta`

~~3926~~

3927 - `MessageDeltaEvent data`

~~3928~~

3929 Represents a message delta i.e. any changed fields on a message during streaming.

~~3930~~

3931 - `String id`

~~3932~~

3933 The identifier of the message, which can be referenced in API endpoints.

~~3934~~

3935 - `MessageDelta delta`

~~3936~~

3937 The delta containing the fields that have changed on the Message.

~~3938~~

3939 - `Optional<List<MessageContentDelta>> content`

~~3940~~

3941 The content of the message in array of text and/or images.

~~3942~~

3943 - `class ImageFileDeltaBlock:`

~~3944~~

3945 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

~~3946~~

3947 - `long index`

~~3948~~

3949 The index of the content part in the message.

~~3950~~

3951 - `JsonValue; type "image_file"constant`

~~3952~~

3953 Always `image_file`.

~~3954~~

3955 - `IMAGE_FILE("image_file")`

~~3956~~

3957 - `Optional<ImageFileDelta> imageFile`

~~3958~~

3959 - `Optional<Detail> detail`

~~3960~~

3961 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`.

~~3962~~

3963 - `AUTO("auto")`

~~3964~~

3965 - `LOW("low")`

~~3966~~

3967 - `HIGH("high")`

~~3968~~

3969 - `Optional<String> fileId`

~~3970~~

3971 The [File](https://platform.openai.com/docs/api-reference/files) ID of the image in the message content. Set `purpose="vision"` when uploading the File if you need to later display the file content.

~~3972~~

3973 - `class TextDeltaBlock:`

~~3974~~

3975 The text content that is part of a message.

~~3976~~

3977 - `long index`

~~3978~~

3979 The index of the content part in the message.

~~3980~~

3981 - `JsonValue; type "text"constant`

~~3982~~

3983 Always `text`.

~~3984~~

3985 - `TEXT("text")`

~~3986~~

3987 - `Optional<TextDelta> text`

~~3988~~

3989 - `Optional<List<AnnotationDelta>> annotations`

~~3990~~

3991 - `class FileCitationDeltaAnnotation:`

~~3992~~

3993 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.

~~3994~~

3995 - `long index`

3996 2152

3997 The index of the annotation in the text content part.2153 - `Optional<Long> completedAt`

3998 2154

3999 - `JsonValue; type "file_citation"constant`2155 The Unix timestamp (in seconds) for when the message was completed.

4000 2156

4001 Always `file_citation`.2157 - `List<JsonValue> content`

4002 2158

4003 - `FILE_CITATION("file_citation")`2159 The content of the message in array of text and/or images.

4004 2160

4005 - `Optional<Long> endIndex`2161 - `long createdAt`

4006 2162

4007 - `Optional<FileCitation> fileCitation`2163 The Unix timestamp (in seconds) for when the message was created.

4008 2164

4009 - `Optional<String> fileId`2165 - `Optional<Long> incompleteAt`

4010 2166

4011 The ID of the specific File the citation is from.2167 The Unix timestamp (in seconds) for when the message was marked as incomplete.

4012 2168

4013 - `Optional<String> quote`2169 - `Optional<IncompleteDetails> incompleteDetails`

4014 2170

4015 The specific quote in the file.2171 On an incomplete message, details about why the message is incomplete.

4016 2172

4017 - `Optional<Long> startIndex`2173 - `Reason reason`

4018 2174

4019 - `Optional<String> text`2175 The reason the message is incomplete.

4020 2176

4021 The text in the message content that needs to be replaced.2177 - `CONTENT_FILTER("content_filter")`

4022 2178

4023 - `class FilePathDeltaAnnotation:`2179 - `MAX_TOKENS("max_tokens")`

4024 2180

4025 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.2181 - `RUN_CANCELLED("run_cancelled")`

4026 2182

4027 - `long index`2183 - `RUN_EXPIRED("run_expired")`

4028 2184

4029 The index of the annotation in the text content part.2185 - `RUN_FAILED("run_failed")`

4030 2186

4031 - `JsonValue; type "file_path"constant`2187 - `Optional<Metadata> metadata`

4032 2188

4033 Always `file_path`.2189 Set of 16 key-value pairs that can be attached to an object. This can be

2190 useful for storing additional information about the object in a structured

2191 format, and querying for objects via API or the dashboard.

4034 2192

4035 - `FILE_PATH("file_path")`2193 Keys are strings with a maximum length of 64 characters. Values are strings

2194 with a maximum length of 512 characters.

4036 2195

4037 - `Optional<Long> endIndex`2196 - `JsonValue; object_ "thread.message"constant`

4038 2197

4039 - `Optional<FilePath> filePath`2198 The object type, which is always `thread.message`.

4040 2199

4041 - `Optional<String> fileId`2200 - `THREAD_MESSAGE("thread.message")`

4042 2201

4043 The ID of the file that was generated.2202 - `Role role`

4044 2203

4045 - `Optional<Long> startIndex`2204 The entity that produced the message. One of `user` or `assistant`.

4046 2205

4047 - `Optional<String> text`2206 - `USER("user")`

4048 2207

4049 The text in the message content that needs to be replaced.2208 - `ASSISTANT("assistant")`

4050 2209

4051 - `Optional<String> value`2210 - `Optional<String> runId`

4052 2211

4053 The data that makes up the text.2212 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) associated with the creation of this message. Value is `null` when messages are created manually using the create message or create thread endpoints.

4054 2213

4055 - `class RefusalDeltaBlock:`2214 - `Status status`

4056 2215

4057 The refusal content that is part of a message.2216 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

4058 2217

4059 - `long index`2218 - `IN_PROGRESS("in_progress")`

4060 2219

4061 The index of the refusal part in the message.2220 - `INCOMPLETE("incomplete")`

4062 2221

4063 - `JsonValue; type "refusal"constant`2222 - `COMPLETED("completed")`

4064 2223

4065 Always `refusal`.2224 - `String threadId`

4066 2225

4067 - `REFUSAL("refusal")`2226 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.

4068 2227

4069 - `Optional<String> refusal`2228 - `JsonValue; event "thread.message.created"constant`

4070 2229

4071 - `class ImageUrlDeltaBlock:`2230 - `THREAD_MESSAGE_CREATED("thread.message.created")`

4072 2231

4073 References an image URL in the content of a message.2232 - `ThreadMessageInProgress`

4074 2233

4075 - `long index`2234 - `Message data`

4076 2235

4077 The index of the content part in the message.2236 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4078 2237

4079 - `JsonValue; type "image_url"constant`2238 - `JsonValue; event "thread.message.in_progress"constant`

4080 2239

4081 Always `image_url`.2240 - `THREAD_MESSAGE_IN_PROGRESS("thread.message.in_progress")`

4082 2241

4083 - `IMAGE_URL("image_url")`2242 - `ThreadMessageDelta`

4084 2243

4085 - `Optional<ImageUrlDelta> imageUrl`2244 - `MessageDeltaEvent data`

4086 2245

4087 - `Optional<Detail> detail`2246 Represents a message delta i.e. any changed fields on a message during streaming.

4088 2247

4089 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`.2248 - `String id`

4090 2249

4091 - `AUTO("auto")`2250 The identifier of the message, which can be referenced in API endpoints.

4092 2251

4093 - `LOW("low")`2252 - `MessageDelta delta`

4094 2253

4095 - `HIGH("high")`2254 The delta containing the fields that have changed on the Message.

4096 2255

4097 - `Optional<String> url`2256 - `Optional<List<JsonValue>> content`

4098 2257

4099 The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.2258 The content of the message in array of text and/or images.

4100 2259

4101 - `Optional<Role> role`2260 - `Optional<Role> role`

4102 2261

4136 2295

4137 - `THREAD_MESSAGE_INCOMPLETE("thread.message.incomplete")`2296 - `THREAD_MESSAGE_INCOMPLETE("thread.message.incomplete")`

4138 2297

4139### Run Step Stream Event2298 - `ErrorEvent`

~~4140~~

4141- `class RunStepStreamEvent: A class that can be one of several variants.union`

~~4142~~

4143 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.

~~4144~~

4145 - `ThreadRunStepCreated`

~~4146~~

4147 - `RunStep data`

4148 2299

4149 Represents a step in execution of a run.2300 - `ErrorObject data`

4150 2301

4151 - `String id`2302 - `Optional<String> code`

4152 2303

4153 The identifier of the run step, which can be referenced in API endpoints.2304 - `String message`

4154 2305

4155 - `String assistantId`2306 - `Optional<String> param`

4156 2307

4157 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) associated with the run step.2308 - `String type`

4158 2309

4159 - `Optional<Long> cancelledAt`2310 - `JsonValue; event "error"constant`

4160 2311

4161 The Unix timestamp (in seconds) for when the run step was cancelled.2312 - `ERROR("error")`

4162 2313

4163 - `Optional<Long> completedAt`2314### Assistant Tool

4164 2315

4165 The Unix timestamp (in seconds) for when the run step completed.2316- `class AssistantTool:`

4166 2317

4167 - `long createdAt`2318### Code Interpreter Tool

4168 2319

4169 The Unix timestamp (in seconds) for when the run step was created.2320- `class CodeInterpreterTool:`

4170 2321

4171 - `Optional<Long> expiredAt`2322 - `JsonValue; type "code_interpreter"constant`

4172 2323

4173 The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.2324 The type of tool being defined: `code_interpreter`

4174 2325

4175 - `Optional<Long> failedAt`2326 - `CODE_INTERPRETER("code_interpreter")`

4176 2327

4177 The Unix timestamp (in seconds) for when the run step failed.2328### File Search Tool

4178 2329

4179 - `Optional<LastError> lastError`2330- `class FileSearchTool:`

4180 2331

4181 The last error associated with this run step. Will be `null` if there are no errors.2332 - `JsonValue; type "file_search"constant`

4182 2333

4183 - `Code code`2334 The type of tool being defined: `file_search`

4184 2335

4185 One of `server_error` or `rate_limit_exceeded`.2336 - `FILE_SEARCH("file_search")`

4186 2337

4187 - `SERVER_ERROR("server_error")`2338 - `Optional<FileSearch> fileSearch`

4188 2339

4189 - `RATE_LIMIT_EXCEEDED("rate_limit_exceeded")`2340 Overrides for the file search tool.

4190 2341

4191 - `String message`2342 - `Optional<Long> maxNumResults`

4192 2343

4193 A human-readable description of the error.2344 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.

4194 2345

4195 - `Optional<Metadata> metadata`2346 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

4196 2347

4197 Set of 16 key-value pairs that can be attached to an object. This can be2348 - `Optional<RankingOptions> rankingOptions`

4198 useful for storing additional information about the object in a structured

4199 format, and querying for objects via API or the dashboard.

4200 2349

4201 Keys are strings with a maximum length of 64 characters. Values are strings2350 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.

4202 with a maximum length of 512 characters.

4203 2351

4204 - `JsonValue; object_ "thread.run.step"constant`2352 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

4205 2353

4206 The object type, which is always `thread.run.step`.2354 - `double scoreThreshold`

4207 2355

4208 - `THREAD_RUN_STEP("thread.run.step")`2356 The score threshold for the file search. All values must be a floating point number between 0 and 1.

4209 2357

4210 - `String runId`2358 - `Optional<Ranker> ranker`

4211 2359

4212 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) that this run step is a part of.2360 The ranker to use for the file search. If not specified will use the `auto` ranker.

4213 2361

4214 - `Status status`2362 - `AUTO("auto")`

4215 2363

4216 The status of the run step, which can be either `in_progress`, `cancelled`, `failed`, `completed`, or `expired`.2364 - `DEFAULT_2024_08_21("default_2024_08_21")`

4217 2365

4218 - `IN_PROGRESS("in_progress")`2366### Function Tool

4219 2367

4220 - `CANCELLED("cancelled")`2368- `class FunctionTool:`

4221 2369

4222 - `FAILED("failed")`2370 - `FunctionDefinition function`

4223 2371

4224 - `COMPLETED("completed")`2372 - `String name`

4225 2373

4226 - `EXPIRED("expired")`2374 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.

4227 2375

4228 - `StepDetails stepDetails`2376 - `Optional<String> description`

4229 2377

4230 The details of the run step.2378 A description of what the function does, used by the model to choose when and how to call the function.

4231 2379

4232 - `class MessageCreationStepDetails:`2380 - `Optional<FunctionParameters> parameters`

4233 2381

4234 Details of the message creation by the run step.2382 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

4235 2383

4236 - `MessageCreation messageCreation`2384 Omitting `parameters` defines a function with an empty parameter list.

4237 2385

4238 - `String messageId`2386 - `Optional<Boolean> strict`

4239 2387

4240 The ID of the message that was created by this run step.2388 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](https://platform.openai.com/docs/guides/function-calling).

4241 2389

4242 - `JsonValue; type "message_creation"constant`2390 - `JsonValue; type "function"constant`

4243 2391

4244 Always `message_creation`.2392 The type of tool being defined: `function`

4245 2393

4246 - `MESSAGE_CREATION("message_creation")`2394 - `FUNCTION("function")`

4247 2395

4248 - `class ToolCallsStepDetails:`2396### Message Stream Event

4249 2397

4250 Details of the tool call.2398- `class MessageStreamEvent: A class that can be one of several variants.union`

4251 2399

4252 - `List<ToolCall> toolCalls`2400 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.

4253 2401

4254 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`.2402 - `ThreadMessageCreated`

4255 2403

4256 - `class CodeInterpreterToolCall:`2404 - `Message data`

4257 2405

4258 Details of the Code Interpreter tool call the run step was involved in.2406 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4259 2407

4260 - `String id`2408 - `String id`

4261 2409

4262 The ID of the tool call.2410 The identifier, which can be referenced in API endpoints.

~~4263~~

4264 - `CodeInterpreter codeInterpreter`

~~4265~~

4266 The Code Interpreter tool call definition.

4267 2411

4268 - `String input`2412 - `Optional<String> assistantId`

4269 2413

4270 The input to the Code Interpreter tool call.2414 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.

4271 2415

4272 - `List<Output> outputs`2416 - `Optional<List<Attachment>> attachments`

4273 2417

4274 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.2418 A list of files attached to the message, and the tools they were added to.

4275 2419

4276 - `class LogsOutput:`2420 - `Optional<String> fileId`

4277 2421

4278 Text output from the Code Interpreter tool call as part of a run step.2422 The ID of the file to attach to the message.

4279 2423

4280 - `String logs`2424 - `Optional<List<Tool>> tools`

4281 2425

4282 The text output from the Code Interpreter tool call.2426 The tools to add this file to.

4283 2427

4284 - `JsonValue; type "logs"constant`2428 - `class CodeInterpreterTool:`

4285 2429

4286 Always `logs`.2430 - `JsonValue; type "code_interpreter"constant`

4287 2431

4288 - `LOGS("logs")`2432 The type of tool being defined: `code_interpreter`

4289 2433

4290 - `class ImageOutput:`2434 - `CODE_INTERPRETER("code_interpreter")`

4291 2435

4292 - `Image image`2436 - `JsonValue;`

4293 2437

4294 - `String fileId`2438 - `JsonValue; type "file_search"constant`

4295 2439

4296 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.2440 The type of tool being defined: `file_search`

4297 2441

4298 - `JsonValue; type "image"constant`2442 - `FILE_SEARCH("file_search")`

4299 2443

4300 Always `image`.2444 - `Optional<Long> completedAt`

4301 2445

4302 - `IMAGE("image")`2446 The Unix timestamp (in seconds) for when the message was completed.

4303 2447

4304 - `JsonValue; type "code_interpreter"constant`2448 - `List<JsonValue> content`

4305 2449

4306 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.2450 The content of the message in array of text and/or images.

4307 2451

4308 - `CODE_INTERPRETER("code_interpreter")`2452 - `long createdAt`

4309 2453

4310 - `class FileSearchToolCall:`2454 The Unix timestamp (in seconds) for when the message was created.

4311 2455

4312 - `String id`2456 - `Optional<Long> incompleteAt`

4313 2457

4314 The ID of the tool call object.2458 The Unix timestamp (in seconds) for when the message was marked as incomplete.

4315 2459

4316 - `FileSearch fileSearch`2460 - `Optional<IncompleteDetails> incompleteDetails`

4317 2461

4318 For now, this is always going to be an empty object.2462 On an incomplete message, details about why the message is incomplete.

4319 2463

4320 - `Optional<RankingOptions> rankingOptions`2464 - `Reason reason`

4321 2465

4322 The ranking options for the file search.2466 The reason the message is incomplete.

4323 2467

4324 - `Ranker ranker`2468 - `CONTENT_FILTER("content_filter")`

4325 2469

4326 The ranker to use for the file search. If not specified will use the `auto` ranker.2470 - `MAX_TOKENS("max_tokens")`

4327 2471

4328 - `AUTO("auto")`2472 - `RUN_CANCELLED("run_cancelled")`

4329 2473

4330 - `DEFAULT_2024_08_21("default_2024_08_21")`2474 - `RUN_EXPIRED("run_expired")`

4331 2475

4332 - `double scoreThreshold`2476 - `RUN_FAILED("run_failed")`

4333 2477

4334 The score threshold for the file search. All values must be a floating point number between 0 and 1.2478 - `Optional<Metadata> metadata`

4335 2479

4336 - `Optional<List<Result>> results`2480 Set of 16 key-value pairs that can be attached to an object. This can be

2481 useful for storing additional information about the object in a structured

2482 format, and querying for objects via API or the dashboard.

4337 2483

4338 The results of the file search.2484 Keys are strings with a maximum length of 64 characters. Values are strings

2485 with a maximum length of 512 characters.

4339 2486

4340 - `String fileId`2487 - `JsonValue; object_ "thread.message"constant`

4341 2488

4342 The ID of the file that result was found in.2489 The object type, which is always `thread.message`.

4343 2490

4344 - `String fileName`2491 - `THREAD_MESSAGE("thread.message")`

4345 2492

4346 The name of the file that result was found in.2493 - `Role role`

4347 2494

4348 - `double score`2495 The entity that produced the message. One of `user` or `assistant`.

4349 2496

4350 The score of the result. All values must be a floating point number between 0 and 1.2497 - `USER("user")`

4351 2498

4352 - `Optional<List<Content>> content`2499 - `ASSISTANT("assistant")`

4353 2500

4354 The content of the result that was found. The content is only included if requested via the include query parameter.2501 - `Optional<String> runId`

4355 2502

4356 - `Optional<String> text`2503 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) associated with the creation of this message. Value is `null` when messages are created manually using the create message or create thread endpoints.

4357 2504

4358 The text content of the file.2505 - `Status status`

4359 2506

4360 - `Optional<Type> type`2507 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

4361 2508

4362 The type of the content.2509 - `IN_PROGRESS("in_progress")`

4363 2510

4364 - `TEXT("text")`2511 - `INCOMPLETE("incomplete")`

4365 2512

4366 - `JsonValue; type "file_search"constant`2513 - `COMPLETED("completed")`

4367 2514

4368 The type of tool call. This is always going to be `file_search` for this type of tool call.2515 - `String threadId`

4369 2516

4370 - `FILE_SEARCH("file_search")`2517 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.

4371 2518

4372 - `class FunctionToolCall:`2519 - `JsonValue; event "thread.message.created"constant`

4373 2520

4374 - `String id`2521 - `THREAD_MESSAGE_CREATED("thread.message.created")`

4375 2522

4376 The ID of the tool call object.2523 - `ThreadMessageInProgress`

4377 2524

4378 - `Function function`2525 - `Message data`

4379 2526

4380 The definition of the function that was called.2527 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4381 2528

4382 - `String arguments`2529 - `JsonValue; event "thread.message.in_progress"constant`

4383 2530

4384 The arguments passed to the function.2531 - `THREAD_MESSAGE_IN_PROGRESS("thread.message.in_progress")`

4385 2532

4386 - `String name`2533 - `ThreadMessageDelta`

4387 2534

4388 The name of the function.2535 - `MessageDeltaEvent data`

4389 2536

4390 - `Optional<String> output`2537 Represents a message delta i.e. any changed fields on a message during streaming.

4391 2538

4392 The output of the function. This will be `null` if the outputs have not been [submitted](https://platform.openai.com/docs/api-reference/runs/submitToolOutputs) yet.2539 - `String id`

4393 2540

4394 - `JsonValue; type "function"constant`2541 The identifier of the message, which can be referenced in API endpoints.

4395 2542

4396 The type of tool call. This is always going to be `function` for this type of tool call.2543 - `MessageDelta delta`

4397 2544

4398 - `FUNCTION("function")`2545 The delta containing the fields that have changed on the Message.

4399 2546

4400 - `JsonValue; type "tool_calls"constant`2547 - `Optional<List<JsonValue>> content`

4401 2548

4402 Always `tool_calls`.2549 The content of the message in array of text and/or images.

4403 2550

4404 - `TOOL_CALLS("tool_calls")`2551 - `Optional<Role> role`

4405 2552

4406 - `String threadId`2553 The entity that produced the message. One of `user` or `assistant`.

4407 2554

4408 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.2555 - `USER("user")`

4409 2556

4410 - `Type type`2557 - `ASSISTANT("assistant")`

4411 2558

4412 The type of run step, which can be either `message_creation` or `tool_calls`.2559 - `JsonValue; object_ "thread.message.delta"constant`

4413 2560

4414 - `MESSAGE_CREATION("message_creation")`2561 The object type, which is always `thread.message.delta`.

4415 2562

4416 - `TOOL_CALLS("tool_calls")`2563 - `THREAD_MESSAGE_DELTA("thread.message.delta")`

4417 2564

4418 - `Optional<Usage> usage`2565 - `JsonValue; event "thread.message.delta"constant`

4419 2566

4420 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.2567 - `THREAD_MESSAGE_DELTA("thread.message.delta")`

4421 2568

4422 - `long completionTokens`2569 - `ThreadMessageCompleted`

4423 2570

4424 Number of completion tokens used over the course of the run step.2571 - `Message data`

4425 2572

4426 - `long promptTokens`2573 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4427 2574

4428 Number of prompt tokens used over the course of the run step.2575 - `JsonValue; event "thread.message.completed"constant`

4429 2576

4430 - `long totalTokens`2577 - `THREAD_MESSAGE_COMPLETED("thread.message.completed")`

4431 2578

4432 Total number of tokens used (prompt + completion).2579 - `ThreadMessageIncomplete`

4433 2580

4434 - `JsonValue; event "thread.run.step.created"constant`2581 - `Message data`

4435 2582

4436 - `THREAD_RUN_STEP_CREATED("thread.run.step.created")`2583 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4437 2584

4438 - `ThreadRunStepInProgress`2585 - `JsonValue; event "thread.message.incomplete"constant`

4439 2586

4440 - `RunStep data`2587 - `THREAD_MESSAGE_INCOMPLETE("thread.message.incomplete")`

4441 2588

4442 Represents a step in execution of a run.2589### Run Step Stream Event

4443 2590

4444 - `JsonValue; event "thread.run.step.in_progress"constant`2591- `class RunStepStreamEvent: A class that can be one of several variants.union`

4445 2592

4446 - `THREAD_RUN_STEP_IN_PROGRESS("thread.run.step.in_progress")`2593 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.

4447 2594

4448 - `ThreadRunStepDelta`2595 - `ThreadRunStepCreated`

4449 2596

4450 - `RunStepDeltaEvent data`2597 - `RunStep data`

4451 2598

4452 Represents a run step delta i.e. any changed fields on a run step during streaming.2599 Represents a step in execution of a run.

4453 2600

4454 - `String id`2601 - `String id`

4455 2602

4456 The identifier of the run step, which can be referenced in API endpoints.2603 The identifier of the run step, which can be referenced in API endpoints.

4457 2604

4458 - `RunStepDelta delta`2605 - `String assistantId`

4459 2606

4460 The delta containing the fields that have changed on the run step.2607 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) associated with the run step.

4461 2608

4462 - `Optional<StepDetails> stepDetails`2609 - `Optional<Long> cancelledAt`

4463 2610

4464 The details of the run step.2611 The Unix timestamp (in seconds) for when the run step was cancelled.

4465 2612

4466 - `class RunStepDeltaMessageDelta:`2613 - `Optional<Long> completedAt`

4467 2614

4468 Details of the message creation by the run step.2615 The Unix timestamp (in seconds) for when the run step completed.

4469 2616

4470 - `JsonValue; type "message_creation"constant`2617 - `long createdAt`

4471 2618

4472 Always `message_creation`.2619 The Unix timestamp (in seconds) for when the run step was created.

4473 2620

4474 - `MESSAGE_CREATION("message_creation")`2621 - `Optional<Long> expiredAt`

4475 2622

4476 - `Optional<MessageCreation> messageCreation`2623 The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.

4477 2624

4478 - `Optional<String> messageId`2625 - `Optional<Long> failedAt`

4479 2626

4480 The ID of the message that was created by this run step.2627 The Unix timestamp (in seconds) for when the run step failed.

4481 2628

4482 - `class ToolCallDeltaObject:`2629 - `Optional<LastError> lastError`

4483 2630

4484 Details of the tool call.2631 The last error associated with this run step. Will be `null` if there are no errors.

4485 2632

4486 - `JsonValue; type "tool_calls"constant`2633 - `Code code`

4487 2634

4488 Always `tool_calls`.2635 One of `server_error` or `rate_limit_exceeded`.

4489 2636

4490 - `TOOL_CALLS("tool_calls")`2637 - `SERVER_ERROR("server_error")`

2638

2639 - `RATE_LIMIT_EXCEEDED("rate_limit_exceeded")`

4491 2640

4492 - `Optional<List<ToolCallDelta>> toolCalls`2641 - `String message`

4493 2642

4494 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`.2643 A human-readable description of the error.

4495 2644

4496 - `class CodeInterpreterToolCallDelta:`2645 - `Optional<Metadata> metadata`

4497 2646

4498 Details of the Code Interpreter tool call the run step was involved in.2647 Set of 16 key-value pairs that can be attached to an object. This can be

2648 useful for storing additional information about the object in a structured

2649 format, and querying for objects via API or the dashboard.

4499 2650

4500 - `long index`2651 Keys are strings with a maximum length of 64 characters. Values are strings

2652 with a maximum length of 512 characters.

4501 2653

4502 The index of the tool call in the tool calls array.2654 - `JsonValue; object_ "thread.run.step"constant`

4503 2655

4504 - `JsonValue; type "code_interpreter"constant`2656 The object type, which is always `thread.run.step`.

4505 2657

4506 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.2658 - `THREAD_RUN_STEP("thread.run.step")`

4507 2659

4508 - `CODE_INTERPRETER("code_interpreter")`2660 - `String runId`

4509 2661

4510 - `Optional<String> id`2662 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) that this run step is a part of.

4511 2663

4512 The ID of the tool call.2664 - `Status status`

4513 2665

4514 - `Optional<CodeInterpreter> codeInterpreter`2666 The status of the run step, which can be either `in_progress`, `cancelled`, `failed`, `completed`, or `expired`.

4515 2667

4516 The Code Interpreter tool call definition.2668 - `IN_PROGRESS("in_progress")`

4517 2669

4518 - `Optional<String> input`2670 - `CANCELLED("cancelled")`

4519 2671

4520 The input to the Code Interpreter tool call.2672 - `FAILED("failed")`

4521 2673

4522 - `Optional<List<Output>> outputs`2674 - `COMPLETED("completed")`

4523 2675

4524 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.2676 - `EXPIRED("expired")`

4525 2677

4526 - `class CodeInterpreterLogs:`2678 - `StepDetails stepDetails`

4527 2679

4528 Text output from the Code Interpreter tool call as part of a run step.2680 The details of the run step.

4529 2681

4530 - `long index`2682 - `class MessageCreationStepDetails:`

4531 2683

4532 The index of the output in the outputs array.2684 Details of the message creation by the run step.

4533 2685

4534 - `JsonValue; type "logs"constant`2686 - `MessageCreation messageCreation`

4535 2687

4536 Always `logs`.2688 - `String messageId`

4537 2689

4538 - `LOGS("logs")`2690 The ID of the message that was created by this run step.

4539 2691

4540 - `Optional<String> logs`2692 - `JsonValue; type "message_creation"constant`

4541 2693

4542 The text output from the Code Interpreter tool call.2694 Always `message_creation`.

4543 2695

4544 - `class CodeInterpreterOutputImage:`2696 - `MESSAGE_CREATION("message_creation")`

4545 2697

4546 - `long index`2698 - `class ToolCallsStepDetails:`

4547 2699

4548 The index of the output in the outputs array.2700 Details of the tool call.

4549 2701

4550 - `JsonValue; type "image"constant`2702 - `List<JsonValue> toolCalls`

4551 2703

4552 Always `image`.2704 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`.

4553 2705

4554 - `IMAGE("image")`2706 - `JsonValue; type "tool_calls"constant`

4555 2707

4556 - `Optional<Image> image`2708 Always `tool_calls`.

4557 2709

4558 - `Optional<String> fileId`2710 - `TOOL_CALLS("tool_calls")`

4559 2711

4560 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.2712 - `String threadId`

4561 2713

4562 - `class FileSearchToolCallDelta:`2714 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.

4563 2715

4564 - `JsonValue fileSearch`2716 - `Type type`

4565 2717

4566 For now, this is always going to be an empty object.2718 The type of run step, which can be either `message_creation` or `tool_calls`.

4567 2719

4568 - `long index`2720 - `MESSAGE_CREATION("message_creation")`

4569 2721

4570 The index of the tool call in the tool calls array.2722 - `TOOL_CALLS("tool_calls")`

4571 2723

4572 - `JsonValue; type "file_search"constant`2724 - `Optional<Usage> usage`

4573 2725

4574 The type of tool call. This is always going to be `file_search` for this type of tool call.2726 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.

4575 2727

4576 - `FILE_SEARCH("file_search")`2728 - `long completionTokens`

4577 2729

4578 - `Optional<String> id`2730 Number of completion tokens used over the course of the run step.

4579 2731

4580 The ID of the tool call object.2732 - `long promptTokens`

4581 2733

4582 - `class FunctionToolCallDelta:`2734 Number of prompt tokens used over the course of the run step.

4583 2735

4584 - `long index`2736 - `long totalTokens`

4585 2737

4586 The index of the tool call in the tool calls array.2738 Total number of tokens used (prompt + completion).

4587 2739

4588 - `JsonValue; type "function"constant`2740 - `JsonValue; event "thread.run.step.created"constant`

4589 2741

4590 The type of tool call. This is always going to be `function` for this type of tool call.2742 - `THREAD_RUN_STEP_CREATED("thread.run.step.created")`

4591 2743

4592 - `FUNCTION("function")`2744 - `ThreadRunStepInProgress`

4593 2745

4594 - `Optional<String> id`2746 - `RunStep data`

4595 2747

4596 The ID of the tool call object.2748 Represents a step in execution of a run.

4597 2749

4598 - `Optional<Function> function`2750 - `JsonValue; event "thread.run.step.in_progress"constant`

4599 2751

4600 The definition of the function that was called.2752 - `THREAD_RUN_STEP_IN_PROGRESS("thread.run.step.in_progress")`

4601 2753

4602 - `Optional<String> arguments`2754 - `ThreadRunStepDelta`

4603 2755

4604 The arguments passed to the function.2756 - `RunStepDeltaEvent data`

4605 2757

4606 - `Optional<String> name`2758 Represents a run step delta i.e. any changed fields on a run step during streaming.

4607 2759

4608 The name of the function.2760 - `String id`

4609 2761

4610 - `Optional<String> output`2762 The identifier of the run step, which can be referenced in API endpoints.

4611 2763

4612 The output of the function. This will be `null` if the outputs have not been [submitted](https://platform.openai.com/docs/api-reference/runs/submitToolOutputs) yet.2764 - `JsonValue delta`

4613 2765

4614 - `JsonValue; object_ "thread.run.step.delta"constant`2766 - `JsonValue; object_ "thread.run.step.delta"constant`

4615 2767

4885 3037

4886 The Unix timestamp (in seconds) for when the run was started.3038 The Unix timestamp (in seconds) for when the run was started.

4887 3039

4888 - `RunStatus status`3040 - `JsonValue status`

~~4889~~

4890 The status of the run, which can be either `queued`, `in_progress`, `requires_action`, `cancelling`, `cancelled`, `failed`, `completed`, `incomplete`, or `expired`.

~~4891~~

4892 - `QUEUED("queued")`

~~4893~~

4894 - `IN_PROGRESS("in_progress")`

~~4895~~

4896 - `REQUIRES_ACTION("requires_action")`

~~4897~~

4898 - `CANCELLING("cancelling")`

~~4899~~

4900 - `CANCELLED("cancelled")`

~~4901~~

4902 - `FAILED("failed")`

~~4903~~

4904 - `COMPLETED("completed")`

~~4905~~

4906 - `INCOMPLETE("incomplete")`

~~4907~~

4908 - `EXPIRED("expired")`

4909 3041

4910 - `String threadId`3042 - `String threadId`

4911 3043

4947 3079

4948 The name of the function to call.3080 The name of the function to call.

4949 3081

4950 - `List<AssistantTool> tools`3082 - `List<JsonValue> tools`

4951 3083

4952 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.3084 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

4953 3085

4954 - `class CodeInterpreterTool:`

~~4955~~

4956 - `JsonValue; type "code_interpreter"constant`

~~4957~~

4958 The type of tool being defined: `code_interpreter`

~~4959~~

4960 - `CODE_INTERPRETER("code_interpreter")`

~~4961~~

4962 - `class FileSearchTool:`

~~4963~~

4964 - `JsonValue; type "file_search"constant`

~~4965~~

4966 The type of tool being defined: `file_search`

~~4967~~

4968 - `FILE_SEARCH("file_search")`

~~4969~~

4970 - `Optional<FileSearch> fileSearch`

~~4971~~

4972 Overrides for the file search tool.

~~4973~~

4974 - `Optional<Long> maxNumResults`

~~4975~~

4976 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.

~~4977~~

4978 Note that the file search tool may output fewer than `max_num_results` results. See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~4979~~

4980 - `Optional<RankingOptions> rankingOptions`

~~4981~~

4982 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.

~~4983~~

4984 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

~~4985~~

4986 - `double scoreThreshold`

~~4987~~

4988 The score threshold for the file search. All values must be a floating point number between 0 and 1.

~~4989~~

4990 - `Optional<Ranker> ranker`

~~4991~~

4992 The ranker to use for the file search. If not specified will use the `auto` ranker.

~~4993~~

4994 - `AUTO("auto")`

~~4995~~

4996 - `DEFAULT_2024_08_21("default_2024_08_21")`

~~4997~~

4998 - `class FunctionTool:`

~~4999~~

5000 - `FunctionDefinition function`

~~5001~~

5002 - `String name`

~~5003~~

5004 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.

~~5005~~

5006 - `Optional<String> description`

~~5007~~

5008 A description of what the function does, used by the model to choose when and how to call the function.

~~5009~~

5010 - `Optional<FunctionParameters> parameters`

~~5011~~

5012 The parameters the functions accepts, described as a JSON Schema object. See the [guide](https://platform.openai.com/docs/guides/function-calling) for examples, and the [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for documentation about the format.

~~5013~~

5014 Omitting `parameters` defines a function with an empty parameter list.

~~5015~~

5016 - `Optional<Boolean> strict`

~~5017~~

5018 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](https://platform.openai.com/docs/guides/function-calling).

~~5019~~

5020 - `JsonValue; type "function"constant`

~~5021~~

5022 The type of tool being defined: `function`

~~5023~~

5024 - `FUNCTION("function")`

~~5025~~

5026 - `Optional<TruncationStrategy> truncationStrategy`3086 - `Optional<TruncationStrategy> truncationStrategy`

5027 3087

5028 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.3088 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

java/resources/beta/subresources/assistants/index.md 2026-07-07 08:02 UTC to 2026-07-09 20:58 UTC

Assistants

List assistants

Parameters

Returns

Example

Response

Create assistant

Parameters

Returns

Example

Response

Retrieve assistant

Parameters

Returns

Example

Response

Modify assistant

Parameters

Returns

Example

Response

Delete assistant

Parameters

Returns

Example

Response

Domain Types

Assistant

Assistant Deleted

Assistant Stream Event

Assistant Tool

Code Interpreter Tool

File Search Tool

Function Tool

Message Stream Event

Run Step Stream Event

Run Stream Event

Thread Stream Event