Assistants

List assistants

beta.assistants.list(AssistantListParams**kwargs) -> SyncCursorPage[Assistant]

get /assistants

Returns a list of assistants.

Parameters

after: Optional[str]

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

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

A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.
order: Optional[Literal["asc", "desc"]]

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

Returns

class Assistant: …

Represents an assistant that can call the model and use tools.
- id: str
  
  The identifier, which can be referenced in API endpoints.
- created_at: int
  
  The Unix timestamp (in seconds) for when the assistant was created.
- description: Optional[str]
  
  The description of the assistant. The maximum length is 512 characters.
- instructions: Optional[str]
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- metadata: Optional[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.
- model: str
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- name: Optional[str]
  
  The name of the assistant. The maximum length is 256 characters.
- object: Literal["assistant"]
  
  The object type, which is always assistant.
  - "assistant"
- tools: List[AssistantTool]
  
  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.
  - class CodeInterpreterTool: …
    - type: Literal["code_interpreter"]
      
      The type of tool being defined: code_interpreter
      - "code_interpreter"
  - class FileSearchTool: …
    - type: Literal["file_search"]
      
      The type of tool being defined: file_search
      - "file_search"
    - file_search: Optional[FileSearch]
      
      Overrides for the file search tool.
      - max_num_results: Optional[int]
        
        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
        
        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
      - ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
        
        See the file search tool documentation for more information.
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
  - class FunctionTool: …
    - function: FunctionDefinition
      - name: str
        
        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the function does, used by the model to choose when and how to call the function.
      - parameters: Optional[FunctionParameters]
        
        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
        
        Omitting parameters defines a function with an empty parameter list.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
    - type: Literal["function"]
      
      The type of tool being defined: function
      - "function"
- response_format: Optional[AssistantResponseFormatOption]
  
  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.
  - Literal["auto"]
    
    auto is the default value
    - "auto"
  - class ResponseFormatText: …
    
    Default response format. Used to generate text responses.
    - type: Literal["text"]
      
      The type of response format being defined. Always 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.
    - type: Literal["json_object"]
      
      The type of response format being defined. Always json_object.
      - "json_object"
  - class ResponseFormatJSONSchema: …
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - json_schema: JSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - name: str
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - schema: Optional[Dict[str, object]]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
    - type: Literal["json_schema"]
      
      The type of response format being defined. Always json_schema.
      - "json_schema"
- temperature: Optional[float]
  
  What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
- tool_resources: Optional[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.
  - code_interpreter: Optional[ToolResourcesCodeInterpreter]
    - file_ids: Optional[List[str]]
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - file_search: Optional[ToolResourcesFileSearch]
    - vector_store_ids: Optional[List[str]]
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- top_p: Optional[float]
  
  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

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),  # This is the default and can be omitted
)
page = client.beta.assistants.list()
page = page.data[0]
print(page.id)

Response

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

Example

from openai import OpenAI
client = OpenAI()

my_assistants = client.beta.assistants.list(
    order="desc",
    limit="20",
)
print(my_assistants.data)

Response

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

Create assistant

beta.assistants.create(AssistantCreateParams**kwargs) -> Assistant

post /assistants

Create an assistant with a model and instructions.

Parameters

model: Union[str, ChatModel]

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.
- str
- Literal["gpt-5.6-sol", "gpt-5.6-terra", "gpt-5.6-luna", 78 more]
  - "gpt-5.6-sol"
  - "gpt-5.6-terra"
  - "gpt-5.6-luna"
  - "gpt-5.4"
  - "gpt-5.4-mini"
  - "gpt-5.4-nano"
  - "gpt-5.4-mini-2026-03-17"
  - "gpt-5.4-nano-2026-03-17"
  - "gpt-5.3-chat-latest"
  - "gpt-5.2"
  - "gpt-5.2-2025-12-11"
  - "gpt-5.2-chat-latest"
  - "gpt-5.2-pro"
  - "gpt-5.2-pro-2025-12-11"
  - "gpt-5.1"
  - "gpt-5.1-2025-11-13"
  - "gpt-5.1-codex"
  - "gpt-5.1-mini"
  - "gpt-5.1-chat-latest"
  - "gpt-5"
  - "gpt-5-mini"
  - "gpt-5-nano"
  - "gpt-5-2025-08-07"
  - "gpt-5-mini-2025-08-07"
  - "gpt-5-nano-2025-08-07"
  - "gpt-5-chat-latest"
  - "gpt-4.1"
  - "gpt-4.1-mini"
  - "gpt-4.1-nano"
  - "gpt-4.1-2025-04-14"
  - "gpt-4.1-mini-2025-04-14"
  - "gpt-4.1-nano-2025-04-14"
  - "o4-mini"
  - "o4-mini-2025-04-16"
  - "o3"
  - "o3-2025-04-16"
  - "o3-mini"
  - "o3-mini-2025-01-31"
  - "o1"
  - "o1-2024-12-17"
  - "o1-preview"
  - "o1-preview-2024-09-12"
  - "o1-mini"
  - "o1-mini-2024-09-12"
  - "gpt-4o"
  - "gpt-4o-2024-11-20"
  - "gpt-4o-2024-08-06"
  - "gpt-4o-2024-05-13"
  - "gpt-4o-audio-preview"
  - "gpt-4o-audio-preview-2024-10-01"
  - "gpt-4o-audio-preview-2024-12-17"
  - "gpt-4o-audio-preview-2025-06-03"
  - "gpt-4o-mini-audio-preview"
  - "gpt-4o-mini-audio-preview-2024-12-17"
  - "gpt-4o-search-preview"
  - "gpt-4o-mini-search-preview"
  - "gpt-4o-search-preview-2025-03-11"
  - "gpt-4o-mini-search-preview-2025-03-11"
  - "chatgpt-4o-latest"
  - "codex-mini-latest"
  - "gpt-4o-mini"
  - "gpt-4o-mini-2024-07-18"
  - "gpt-4-turbo"
  - "gpt-4-turbo-2024-04-09"
  - "gpt-4-0125-preview"
  - "gpt-4-turbo-preview"
  - "gpt-4-1106-preview"
  - "gpt-4-vision-preview"
  - "gpt-4"
  - "gpt-4-0314"
  - "gpt-4-0613"
  - "gpt-4-32k"
  - "gpt-4-32k-0314"
  - "gpt-4-32k-0613"
  - "gpt-3.5-turbo"
  - "gpt-3.5-turbo-16k"
  - "gpt-3.5-turbo-0301"
  - "gpt-3.5-turbo-0613"
  - "gpt-3.5-turbo-1106"
  - "gpt-3.5-turbo-0125"
  - "gpt-3.5-turbo-16k-0613"
description: Optional[str]

The description of the assistant. The maximum length is 512 characters.
instructions: Optional[str]

The system instructions that the assistant uses. The maximum length is 256,000 characters.
metadata: Optional[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.
name: Optional[str]

The name of the assistant. The maximum length is 256 characters.
reasoning_effort: Optional[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.
- "none"
- "minimal"
- "low"
- "medium"
- "high"
- "xhigh"
- "max"
response_format: Optional[AssistantResponseFormatOptionParam]

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.
- Literal["auto"]
  
  auto is the default value
  - "auto"
- class ResponseFormatText: …
  
  Default response format. Used to generate text responses.
  - type: Literal["text"]
    
    The type of response format being defined. Always 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.
  - type: Literal["json_object"]
    
    The type of response format being defined. Always json_object.
    - "json_object"
- class ResponseFormatJSONSchema: …
  
  JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
  - json_schema: JSONSchema
    
    Structured Outputs configuration options, including a JSON Schema.
    - name: str
      
      The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
    - description: Optional[str]
      
      A description of what the response format is for, used by the model to determine how to respond in the format.
    - schema: Optional[Dict[str, object]]
      
      The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
    - strict: Optional[bool]
      
      Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
  - type: Literal["json_schema"]
    
    The type of response format being defined. Always json_schema.
    - "json_schema"
temperature: Optional[float]

What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
tool_resources: Optional[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.
- code_interpreter: Optional[ToolResourcesCodeInterpreter]
  - file_ids: Optional[Sequence[str]]
    
    A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
- file_search: Optional[ToolResourcesFileSearch]
  - vector_store_ids: Optional[Sequence[str]]
    
    The vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
  - vector_stores: Optional[Iterable[ToolResourcesFileSearchVectorStore]]
    
    A helper to create a vector store with file_ids and attach it to this assistant. There can be a maximum of 1 vector store attached to the assistant.
    - chunking_strategy: Optional[ToolResourcesFileSearchVectorStoreChunkingStrategy]
      
      The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.
      - class ToolResourcesFileSearchVectorStoreChunkingStrategyAuto: …
        
        The default strategy. This strategy currently uses a max_chunk_size_tokens of 800 and chunk_overlap_tokens of 400.
        
        type: Literal["auto"]
        
        Always auto.
        
        "auto"
      - class ToolResourcesFileSearchVectorStoreChunkingStrategyStatic: …
        
        static: ToolResourcesFileSearchVectorStoreChunkingStrategyStaticStatic
        
        chunk_overlap_tokens: int
        
        The number of tokens that overlap between chunks. The default value is 400.
        
        Note that the overlap must not exceed half of max_chunk_size_tokens.
        
        max_chunk_size_tokens: int
        
        The maximum number of tokens in each chunk. The default value is 800. The minimum value is 100 and the maximum value is 4096.
        
        type: Literal["static"]
        
        Always static.
        
        "static"
    - file_ids: Optional[Sequence[str]]
      
      A list of file IDs to add to the vector store. For vector stores created before Nov 2025, there can be a maximum of 10,000 files in a vector store. For vector stores created starting in Nov 2025, the limit is 100,000,000 files.
    - metadata: Optional[Metadata]
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
tools: Optional[Iterable[AssistantToolParam]]

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.
- class CodeInterpreterTool: …
  - type: Literal["code_interpreter"]
    
    The type of tool being defined: code_interpreter
    - "code_interpreter"
- class FileSearchTool: …
  - type: Literal["file_search"]
    
    The type of tool being defined: file_search
    - "file_search"
  - file_search: Optional[FileSearch]
    
    Overrides for the file search tool.
    - max_num_results: Optional[int]
      
      The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
      
      Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
    - ranking_options: Optional[FileSearchRankingOptions]
      
      The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
      
      See the file search tool documentation for more information.
      - score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
      - ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
- class FunctionTool: …
  - function: FunctionDefinition
    - name: str
      
      The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
    - description: Optional[str]
      
      A description of what the function does, used by the model to choose when and how to call the function.
    - parameters: Optional[FunctionParameters]
      
      The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
      
      Omitting parameters defines a function with an empty parameter list.
    - strict: Optional[bool]
      
      Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
  - type: Literal["function"]
    
    The type of tool being defined: function
    - "function"
top_p: Optional[float]

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.
- id: str
  
  The identifier, which can be referenced in API endpoints.
- created_at: int
  
  The Unix timestamp (in seconds) for when the assistant was created.
- description: Optional[str]
  
  The description of the assistant. The maximum length is 512 characters.
- instructions: Optional[str]
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- metadata: Optional[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.
- model: str
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- name: Optional[str]
  
  The name of the assistant. The maximum length is 256 characters.
- object: Literal["assistant"]
  
  The object type, which is always assistant.
  - "assistant"
- tools: List[AssistantTool]
  
  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.
  - class CodeInterpreterTool: …
    - type: Literal["code_interpreter"]
      
      The type of tool being defined: code_interpreter
      - "code_interpreter"
  - class FileSearchTool: …
    - type: Literal["file_search"]
      
      The type of tool being defined: file_search
      - "file_search"
    - file_search: Optional[FileSearch]
      
      Overrides for the file search tool.
      - max_num_results: Optional[int]
        
        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
        
        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
      - ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
        
        See the file search tool documentation for more information.
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
  - class FunctionTool: …
    - function: FunctionDefinition
      - name: str
        
        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the function does, used by the model to choose when and how to call the function.
      - parameters: Optional[FunctionParameters]
        
        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
        
        Omitting parameters defines a function with an empty parameter list.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
    - type: Literal["function"]
      
      The type of tool being defined: function
      - "function"
- response_format: Optional[AssistantResponseFormatOption]
  
  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.
  - Literal["auto"]
    
    auto is the default value
    - "auto"
  - class ResponseFormatText: …
    
    Default response format. Used to generate text responses.
    - type: Literal["text"]
      
      The type of response format being defined. Always 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.
    - type: Literal["json_object"]
      
      The type of response format being defined. Always json_object.
      - "json_object"
  - class ResponseFormatJSONSchema: …
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - json_schema: JSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - name: str
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - schema: Optional[Dict[str, object]]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
    - type: Literal["json_schema"]
      
      The type of response format being defined. Always json_schema.
      - "json_schema"
- temperature: Optional[float]
  
  What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
- tool_resources: Optional[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.
  - code_interpreter: Optional[ToolResourcesCodeInterpreter]
    - file_ids: Optional[List[str]]
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - file_search: Optional[ToolResourcesFileSearch]
    - vector_store_ids: Optional[List[str]]
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- top_p: Optional[float]
  
  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

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),  # This is the default and can be omitted
)
assistant = client.beta.assistants.create(
    model="gpt-4o",
)
print(assistant.id)

Response

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

Code Interpreter

from openai import OpenAI
client = OpenAI()

my_assistant = client.beta.assistants.create(
    instructions="You are a personal math tutor. When asked a question, write and run Python code to answer the question.",
    name="Math Tutor",
    tools=[{"type": "code_interpreter"}],
    model="gpt-4o",
)
print(my_assistant)

Response

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

Files

from openai import OpenAI
client = OpenAI()

my_assistant = client.beta.assistants.create(
    instructions="You are an HR bot, and you have access to files to answer employee questions about company policies.",
    name="HR Helper",
    tools=[{"type": "file_search"}],
    tool_resources={"file_search": {"vector_store_ids": ["vs_123"]}},
    model="gpt-4o"
)
print(my_assistant)

Response

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

Retrieve assistant

beta.assistants.retrieve(strassistant_id) -> Assistant

get /assistants/{assistant_id}

Retrieves an assistant.

Parameters

assistant_id: str

Returns

class Assistant: …

Represents an assistant that can call the model and use tools.
- id: str
  
  The identifier, which can be referenced in API endpoints.
- created_at: int
  
  The Unix timestamp (in seconds) for when the assistant was created.
- description: Optional[str]
  
  The description of the assistant. The maximum length is 512 characters.
- instructions: Optional[str]
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- metadata: Optional[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.
- model: str
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- name: Optional[str]
  
  The name of the assistant. The maximum length is 256 characters.
- object: Literal["assistant"]
  
  The object type, which is always assistant.
  - "assistant"
- tools: List[AssistantTool]
  
  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.
  - class CodeInterpreterTool: …
    - type: Literal["code_interpreter"]
      
      The type of tool being defined: code_interpreter
      - "code_interpreter"
  - class FileSearchTool: …
    - type: Literal["file_search"]
      
      The type of tool being defined: file_search
      - "file_search"
    - file_search: Optional[FileSearch]
      
      Overrides for the file search tool.
      - max_num_results: Optional[int]
        
        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
        
        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
      - ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
        
        See the file search tool documentation for more information.
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
  - class FunctionTool: …
    - function: FunctionDefinition
      - name: str
        
        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the function does, used by the model to choose when and how to call the function.
      - parameters: Optional[FunctionParameters]
        
        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
        
        Omitting parameters defines a function with an empty parameter list.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
    - type: Literal["function"]
      
      The type of tool being defined: function
      - "function"
- response_format: Optional[AssistantResponseFormatOption]
  
  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.
  - Literal["auto"]
    
    auto is the default value
    - "auto"
  - class ResponseFormatText: …
    
    Default response format. Used to generate text responses.
    - type: Literal["text"]
      
      The type of response format being defined. Always 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.
    - type: Literal["json_object"]
      
      The type of response format being defined. Always json_object.
      - "json_object"
  - class ResponseFormatJSONSchema: …
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - json_schema: JSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - name: str
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - schema: Optional[Dict[str, object]]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
    - type: Literal["json_schema"]
      
      The type of response format being defined. Always json_schema.
      - "json_schema"
- temperature: Optional[float]
  
  What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
- tool_resources: Optional[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.
  - code_interpreter: Optional[ToolResourcesCodeInterpreter]
    - file_ids: Optional[List[str]]
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - file_search: Optional[ToolResourcesFileSearch]
    - vector_store_ids: Optional[List[str]]
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- top_p: Optional[float]
  
  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

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),  # This is the default and can be omitted
)
assistant = client.beta.assistants.retrieve(
    "assistant_id",
)
print(assistant.id)

Response

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

Example

from openai import OpenAI
client = OpenAI()

my_assistant = client.beta.assistants.retrieve("asst_abc123")
print(my_assistant)

Response

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

Modify assistant

beta.assistants.update(strassistant_id, AssistantUpdateParams**kwargs) -> Assistant

post /assistants/{assistant_id}

Modifies an assistant.

Parameters

assistant_id: str
description: Optional[str]

The description of the assistant. The maximum length is 512 characters.
instructions: Optional[str]

The system instructions that the assistant uses. The maximum length is 256,000 characters.
metadata: Optional[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.
model: Optional[Union[str, Literal["gpt-5", "gpt-5-mini", "gpt-5-nano", 39 more]]]

ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- str
- Literal["gpt-5", "gpt-5-mini", "gpt-5-nano", 39 more]
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
  - "gpt-5"
  - "gpt-5-mini"
  - "gpt-5-nano"
  - "gpt-5-2025-08-07"
  - "gpt-5-mini-2025-08-07"
  - "gpt-5-nano-2025-08-07"
  - "gpt-4.1"
  - "gpt-4.1-mini"
  - "gpt-4.1-nano"
  - "gpt-4.1-2025-04-14"
  - "gpt-4.1-mini-2025-04-14"
  - "gpt-4.1-nano-2025-04-14"
  - "o3-mini"
  - "o3-mini-2025-01-31"
  - "o1"
  - "o1-2024-12-17"
  - "gpt-4o"
  - "gpt-4o-2024-11-20"
  - "gpt-4o-2024-08-06"
  - "gpt-4o-2024-05-13"
  - "gpt-4o-mini"
  - "gpt-4o-mini-2024-07-18"
  - "gpt-4.5-preview"
  - "gpt-4.5-preview-2025-02-27"
  - "gpt-4-turbo"
  - "gpt-4-turbo-2024-04-09"
  - "gpt-4-0125-preview"
  - "gpt-4-turbo-preview"
  - "gpt-4-1106-preview"
  - "gpt-4-vision-preview"
  - "gpt-4"
  - "gpt-4-0314"
  - "gpt-4-0613"
  - "gpt-4-32k"
  - "gpt-4-32k-0314"
  - "gpt-4-32k-0613"
  - "gpt-3.5-turbo"
  - "gpt-3.5-turbo-16k"
  - "gpt-3.5-turbo-0613"
  - "gpt-3.5-turbo-1106"
  - "gpt-3.5-turbo-0125"
  - "gpt-3.5-turbo-16k-0613"
name: Optional[str]

The name of the assistant. The maximum length is 256 characters.
reasoning_effort: Optional[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.
- "none"
- "minimal"
- "low"
- "medium"
- "high"
- "xhigh"
- "max"
response_format: Optional[AssistantResponseFormatOptionParam]

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.
- Literal["auto"]
  
  auto is the default value
  - "auto"
- class ResponseFormatText: …
  
  Default response format. Used to generate text responses.
  - type: Literal["text"]
    
    The type of response format being defined. Always 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.
  - type: Literal["json_object"]
    
    The type of response format being defined. Always json_object.
    - "json_object"
- class ResponseFormatJSONSchema: …
  
  JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
  - json_schema: JSONSchema
    
    Structured Outputs configuration options, including a JSON Schema.
    - name: str
      
      The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
    - description: Optional[str]
      
      A description of what the response format is for, used by the model to determine how to respond in the format.
    - schema: Optional[Dict[str, object]]
      
      The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
    - strict: Optional[bool]
      
      Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
  - type: Literal["json_schema"]
    
    The type of response format being defined. Always json_schema.
    - "json_schema"
temperature: Optional[float]

What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
tool_resources: Optional[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.
- code_interpreter: Optional[ToolResourcesCodeInterpreter]
  - file_ids: Optional[Sequence[str]]
    
    Overrides the list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
- file_search: Optional[ToolResourcesFileSearch]
  - vector_store_ids: Optional[Sequence[str]]
    
    Overrides the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
tools: Optional[Iterable[AssistantToolParam]]

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.
- class CodeInterpreterTool: …
  - type: Literal["code_interpreter"]
    
    The type of tool being defined: code_interpreter
    - "code_interpreter"
- class FileSearchTool: …
  - type: Literal["file_search"]
    
    The type of tool being defined: file_search
    - "file_search"
  - file_search: Optional[FileSearch]
    
    Overrides for the file search tool.
    - max_num_results: Optional[int]
      
      The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
      
      Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
    - ranking_options: Optional[FileSearchRankingOptions]
      
      The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
      
      See the file search tool documentation for more information.
      - score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
      - ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
- class FunctionTool: …
  - function: FunctionDefinition
    - name: str
      
      The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
    - description: Optional[str]
      
      A description of what the function does, used by the model to choose when and how to call the function.
    - parameters: Optional[FunctionParameters]
      
      The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
      
      Omitting parameters defines a function with an empty parameter list.
    - strict: Optional[bool]
      
      Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
  - type: Literal["function"]
    
    The type of tool being defined: function
    - "function"
top_p: Optional[float]

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.
- id: str
  
  The identifier, which can be referenced in API endpoints.
- created_at: int
  
  The Unix timestamp (in seconds) for when the assistant was created.
- description: Optional[str]
  
  The description of the assistant. The maximum length is 512 characters.
- instructions: Optional[str]
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- metadata: Optional[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.
- model: str
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- name: Optional[str]
  
  The name of the assistant. The maximum length is 256 characters.
- object: Literal["assistant"]
  
  The object type, which is always assistant.
  - "assistant"
- tools: List[AssistantTool]
  
  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.
  - class CodeInterpreterTool: …
    - type: Literal["code_interpreter"]
      
      The type of tool being defined: code_interpreter
      - "code_interpreter"
  - class FileSearchTool: …
    - type: Literal["file_search"]
      
      The type of tool being defined: file_search
      - "file_search"
    - file_search: Optional[FileSearch]
      
      Overrides for the file search tool.
      - max_num_results: Optional[int]
        
        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
        
        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
      - ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
        
        See the file search tool documentation for more information.
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
  - class FunctionTool: …
    - function: FunctionDefinition
      - name: str
        
        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the function does, used by the model to choose when and how to call the function.
      - parameters: Optional[FunctionParameters]
        
        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
        
        Omitting parameters defines a function with an empty parameter list.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
    - type: Literal["function"]
      
      The type of tool being defined: function
      - "function"
- response_format: Optional[AssistantResponseFormatOption]
  
  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.
  - Literal["auto"]
    
    auto is the default value
    - "auto"
  - class ResponseFormatText: …
    
    Default response format. Used to generate text responses.
    - type: Literal["text"]
      
      The type of response format being defined. Always 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.
    - type: Literal["json_object"]
      
      The type of response format being defined. Always json_object.
      - "json_object"
  - class ResponseFormatJSONSchema: …
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - json_schema: JSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - name: str
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - schema: Optional[Dict[str, object]]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
    - type: Literal["json_schema"]
      
      The type of response format being defined. Always json_schema.
      - "json_schema"
- temperature: Optional[float]
  
  What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
- tool_resources: Optional[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.
  - code_interpreter: Optional[ToolResourcesCodeInterpreter]
    - file_ids: Optional[List[str]]
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - file_search: Optional[ToolResourcesFileSearch]
    - vector_store_ids: Optional[List[str]]
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- top_p: Optional[float]
  
  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

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),  # This is the default and can be omitted
)
assistant = client.beta.assistants.update(
    assistant_id="assistant_id",
)
print(assistant.id)

Response

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

Example

from openai import OpenAI
client = OpenAI()

my_updated_assistant = client.beta.assistants.update(
  "asst_abc123",
  instructions="You are an HR bot, and you have access to files to answer employee questions about company policies. Always response with info from either of the files.",
  name="HR Helper",
  tools=[{"type": "file_search"}],
  model="gpt-4o"
)

print(my_updated_assistant)

Response

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

Delete assistant

beta.assistants.delete(strassistant_id) -> AssistantDeleted

delete /assistants/{assistant_id}

Delete an assistant.

Parameters

assistant_id: str

Returns

class AssistantDeleted: …
- id: str
- deleted: bool
- object: Literal["assistant.deleted"]
  - "assistant.deleted"

Example

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("OPENAI_API_KEY"),  # This is the default and can be omitted
)
assistant_deleted = client.beta.assistants.delete(
    "assistant_id",
)
print(assistant_deleted.id)

Response

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

Example

from openai import OpenAI
client = OpenAI()

response = client.beta.assistants.delete("asst_abc123")
print(response)

Response

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

Domain Types

Assistant

class Assistant: …

Represents an assistant that can call the model and use tools.
- id: str
  
  The identifier, which can be referenced in API endpoints.
- created_at: int
  
  The Unix timestamp (in seconds) for when the assistant was created.
- description: Optional[str]
  
  The description of the assistant. The maximum length is 512 characters.
- instructions: Optional[str]
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- metadata: Optional[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.
- model: str
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- name: Optional[str]
  
  The name of the assistant. The maximum length is 256 characters.
- object: Literal["assistant"]
  
  The object type, which is always assistant.
  - "assistant"
- tools: List[AssistantTool]
  
  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.
  - class CodeInterpreterTool: …
    - type: Literal["code_interpreter"]
      
      The type of tool being defined: code_interpreter
      - "code_interpreter"
  - class FileSearchTool: …
    - type: Literal["file_search"]
      
      The type of tool being defined: file_search
      - "file_search"
    - file_search: Optional[FileSearch]
      
      Overrides for the file search tool.
      - max_num_results: Optional[int]
        
        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
        
        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
      - ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
        
        See the file search tool documentation for more information.
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
  - class FunctionTool: …
    - function: FunctionDefinition
      - name: str
        
        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the function does, used by the model to choose when and how to call the function.
      - parameters: Optional[FunctionParameters]
        
        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
        
        Omitting parameters defines a function with an empty parameter list.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
    - type: Literal["function"]
      
      The type of tool being defined: function
      - "function"
- response_format: Optional[AssistantResponseFormatOption]
  
  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.
  - Literal["auto"]
    
    auto is the default value
    - "auto"
  - class ResponseFormatText: …
    
    Default response format. Used to generate text responses.
    - type: Literal["text"]
      
      The type of response format being defined. Always 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.
    - type: Literal["json_object"]
      
      The type of response format being defined. Always json_object.
      - "json_object"
  - class ResponseFormatJSONSchema: …
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - json_schema: JSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - name: str
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - description: Optional[str]
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - schema: Optional[Dict[str, object]]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
    - type: Literal["json_schema"]
      
      The type of response format being defined. Always json_schema.
      - "json_schema"
- temperature: Optional[float]
  
  What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
- tool_resources: Optional[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.
  - code_interpreter: Optional[ToolResourcesCodeInterpreter]
    - file_ids: Optional[List[str]]
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - file_search: Optional[ToolResourcesFileSearch]
    - vector_store_ids: Optional[List[str]]
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- top_p: Optional[float]
  
  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: …
- id: str
- deleted: bool
- object: Literal["assistant.deleted"]
  - "assistant.deleted"

Assistant Stream Event

AssistantStreamEvent

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.
- class ThreadCreated: …
  
  Occurs when a new thread is created.
  - data: Thread
    
    Represents a thread that contains messages.
    - id: str
      
      The identifier, which can be referenced in API endpoints.
    - created_at: int
      
      The Unix timestamp (in seconds) for when the thread was created.
    - metadata: Optional[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.
    - object: Literal["thread"]
      
      The object type, which is always thread.
      - "thread"
    - tool_resources: Optional[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.
      - code_interpreter: Optional[ToolResourcesCodeInterpreter]
        
        file_ids: Optional[List[str]]
        
        A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
      - file_search: Optional[ToolResourcesFileSearch]
        
        vector_store_ids: Optional[List[str]]
        
        The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.
  - event: Literal["thread.created"]
    - "thread.created"
  - enabled: Optional[bool]
    
    Whether to enable input audio transcription.
- class ThreadRunCreated: …
  
  Occurs when a new run is created.
  - data: Run
    
    Represents an execution run on a thread.
    - id: str
      
      The identifier, which can be referenced in API endpoints.
    - assistant_id: str
      
      The ID of the assistant used for execution of this run.
    - cancelled_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run was cancelled.
    - completed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run was completed.
    - created_at: int
      
      The Unix timestamp (in seconds) for when the run was created.
    - expires_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run will expire.
    - failed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run failed.
    - incomplete_details: Optional[IncompleteDetails]
      
      Details on why the run is incomplete. Will be null if the run is not incomplete.
      - reason: Optional[Literal["max_completion_tokens", "max_prompt_tokens"]]
        
        The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.
        
        "max_completion_tokens"
        
        "max_prompt_tokens"
    - instructions: str
      
      The instructions that the assistant used for this run.
    - last_error: Optional[LastError]
      
      The last error associated with this run. Will be null if there are no errors.
      - code: Literal["server_error", "rate_limit_exceeded", "invalid_prompt"]
        
        One of server_error, rate_limit_exceeded, or invalid_prompt.
        
        "server_error"
        
        "rate_limit_exceeded"
        
        "invalid_prompt"
      - message: str
        
        A human-readable description of the error.
    - max_completion_tokens: Optional[int]
      
      The maximum number of completion tokens specified to have been used over the course of the run.
    - max_prompt_tokens: Optional[int]
      
      The maximum number of prompt tokens specified to have been used over the course of the run.
    - metadata: Optional[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.
    - model: str
      
      The model that the assistant used for this run.
    - object: Literal["thread.run"]
      
      The object type, which is always thread.run.
      - "thread.run"
    - parallel_tool_calls: bool
      
      Whether to enable parallel function calling during tool use.
    - required_action: Optional[RequiredAction]
      
      Details on the action required to continue the run. Will be null if no action is required.
      - submit_tool_outputs: RequiredActionSubmitToolOutputs
        
        Details on the tool outputs needed for this run to continue.
        
        tool_calls: List[RequiredActionFunctionToolCall]
        
        A list of the relevant tool calls.
        
        id: str
        
        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.
        
        arguments: str
        
        The arguments that the model expects you to pass to the function.
        
        name: str
        
        The name of the function.
        
        type: Literal["function"]
        
        The type of tool call the output is required for. For now, this is always function.
        
        "function"
      - type: Literal["submit_tool_outputs"]
        
        For now, this is always submit_tool_outputs.
        
        "submit_tool_outputs"
    - response_format: Optional[AssistantResponseFormatOption]
      
      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.
      - Literal["auto"]
        
        auto is the default value
        
        "auto"
      - class ResponseFormatText: …
        
        Default response format. Used to generate text responses.
        
        type: Literal["text"]
        
        The type of response format being defined. Always 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.
        
        type: Literal["json_object"]
        
        The type of response format being defined. Always json_object.
        
        "json_object"
      - class ResponseFormatJSONSchema: …
        
        JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
        
        json_schema: JSONSchema
        
        Structured Outputs configuration options, including a JSON Schema.
        
        name: str
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
        
        description: Optional[str]
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
        
        schema: Optional[Dict[str, object]]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
        
        strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
        
        type: Literal["json_schema"]
        
        The type of response format being defined. Always json_schema.
        
        "json_schema"
    - started_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run was started.
    - status: RunStatus
      
      The status of the run, which can be either queued, in_progress, requires_action, cancelling, cancelled, failed, completed, incomplete, or expired.
      - "queued"
      - "in_progress"
      - "requires_action"
      - "cancelling"
      - "cancelled"
      - "failed"
      - "completed"
      - "incomplete"
      - "expired"
    - thread_id: str
      
      The ID of the thread that was executed on as a part of this run.
    - tool_choice: Optional[AssistantToolChoiceOption]
      
      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.
      - Literal["none", "auto", "required"]
        
        none means the model will not call any tools and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools before responding to the user.
        
        "none"
        
        "auto"
        
        "required"
      - class AssistantToolChoice: …
        
        Specifies a tool the model should use. Use to force the model to call a specific tool.
        
        type: Literal["function", "code_interpreter", "file_search"]
        
        The type of the tool. If type is function, the function name must be set
        
        "function"
        
        "code_interpreter"
        
        "file_search"
        
        function: Optional[AssistantToolChoiceFunction]
        
        name: str
        
        The name of the function to call.
    - tools: List[AssistantTool]
      
      The list of tools that the assistant used for this run.
      - class CodeInterpreterTool: …
        
        type: Literal["code_interpreter"]
        
        The type of tool being defined: code_interpreter
        
        "code_interpreter"
      - class FileSearchTool: …
        
        type: Literal["file_search"]
        
        The type of tool being defined: file_search
        
        "file_search"
        
        file_search: Optional[FileSearch]
        
        Overrides for the file search tool.
        
        max_num_results: Optional[int]
        
        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
        
        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
        
        ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
        
        See the file search tool documentation for more information.
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
      - class FunctionTool: …
        
        function: FunctionDefinition
        
        name: str
        
        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
        
        description: Optional[str]
        
        A description of what the function does, used by the model to choose when and how to call the function.
        
        parameters: Optional[FunctionParameters]
        
        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
        
        Omitting parameters defines a function with an empty parameter list.
        
        strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
        
        type: Literal["function"]
        
        The type of tool being defined: function
        
        "function"
    - truncation_strategy: Optional[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: Literal["auto", "last_messages"]
        
        The truncation strategy to use for the thread. The default is auto. If set to last_messages, the thread will be truncated to the n most recent messages in the thread. When set to auto, messages in the middle of the thread will be dropped to fit the context length of the model, max_prompt_tokens.
        
        "auto"
        
        "last_messages"
      - last_messages: Optional[int]
        
        The number of most recent messages from the thread when constructing the context for the run.
    - usage: Optional[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.).
      - completion_tokens: int
        
        Number of completion tokens used over the course of the run.
      - prompt_tokens: int
        
        Number of prompt tokens used over the course of the run.
      - total_tokens: int
        
        Total number of tokens used (prompt + completion).
    - temperature: Optional[float]
      
      The sampling temperature used for this run. If not set, defaults to 1.
    - top_p: Optional[float]
      
      The nucleus sampling value used for this run. If not set, defaults to 1.
  - event: Literal["thread.run.created"]
    - "thread.run.created"
- class ThreadRunQueued: …
  
  Occurs when a run moves to a queued status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.queued"]
    - "thread.run.queued"
- class ThreadRunInProgress: …
  
  Occurs when a run moves to an in_progress status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.in_progress"]
    - "thread.run.in_progress"
- class ThreadRunRequiresAction: …
  
  Occurs when a run moves to a requires_action status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.requires_action"]
    - "thread.run.requires_action"
- class ThreadRunCompleted: …
  
  Occurs when a run is completed.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.completed"]
    - "thread.run.completed"
- class ThreadRunIncomplete: …
  
  Occurs when a run ends with status incomplete.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.incomplete"]
    - "thread.run.incomplete"
- class ThreadRunFailed: …
  
  Occurs when a run fails.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.failed"]
    - "thread.run.failed"
- class ThreadRunCancelling: …
  
  Occurs when a run moves to a cancelling status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.cancelling"]
    - "thread.run.cancelling"
- class ThreadRunCancelled: …
  
  Occurs when a run is cancelled.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.cancelled"]
    - "thread.run.cancelled"
- class ThreadRunExpired: …
  
  Occurs when a run expires.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.expired"]
    - "thread.run.expired"
- class ThreadRunStepCreated: …
  
  Occurs when a run step is created.
  - data: RunStep
    
    Represents a step in execution of a run.
    - id: str
      
      The identifier of the run step, which can be referenced in API endpoints.
    - assistant_id: str
      
      The ID of the assistant associated with the run step.
    - cancelled_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step was cancelled.
    - completed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step completed.
    - created_at: int
      
      The Unix timestamp (in seconds) for when the run step was created.
    - expired_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.
    - failed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step failed.
    - last_error: Optional[LastError]
      
      The last error associated with this run step. Will be null if there are no errors.
      - code: Literal["server_error", "rate_limit_exceeded"]
        
        One of server_error or rate_limit_exceeded.
        
        "server_error"
        
        "rate_limit_exceeded"
      - message: str
        
        A human-readable description of the error.
    - metadata: Optional[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.
    - object: Literal["thread.run.step"]
      
      The object type, which is always thread.run.step.
      - "thread.run.step"
    - run_id: str
      
      The ID of the run that this run step is a part of.
    - status: Literal["in_progress", "cancelled", "failed", 2 more]
      
      The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.
      - "in_progress"
      - "cancelled"
      - "failed"
      - "completed"
      - "expired"
    - step_details: StepDetails
      
      The details of the run step.
      - class MessageCreationStepDetails: …
        
        Details of the message creation by the run step.
        
        message_creation: MessageCreation
        
        message_id: str
        
        The ID of the message that was created by this run step.
        
        type: Literal["message_creation"]
        
        Always message_creation.
        
        "message_creation"
      - class ToolCallsStepDetails: …
        
        Details of the tool call.
        
        tool_calls: List[ToolCall]
        
        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.
        
        class CodeInterpreterToolCall: …
        
        Details of the Code Interpreter tool call the run step was involved in.
        
        id: str
        
        The ID of the tool call.
        
        code_interpreter: CodeInterpreter
        
        The Code Interpreter tool call definition.
        
        input: str
        
        The input to the Code Interpreter tool call.
        
        outputs: List[CodeInterpreterOutput]
        
        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.
        
        class CodeInterpreterOutputLogs: …
        
        Text output from the Code Interpreter tool call as part of a run step.
        
        logs: str
        
        The text output from the Code Interpreter tool call.
        
        type: Literal["logs"]
        
        Always logs.
        
        "logs"
        
        class CodeInterpreterOutputImage: …
        
        image: CodeInterpreterOutputImageImage
        
        file_id: str
        
        The file ID of the image.
        
        type: Literal["image"]
        
        Always image.
        
        "image"
        
        type: Literal["code_interpreter"]
        
        The type of tool call. This is always going to be code_interpreter for this type of tool call.
        
        "code_interpreter"
        
        class FileSearchToolCall: …
        
        id: str
        
        The ID of the tool call object.
        
        file_search: FileSearch
        
        For now, this is always going to be an empty object.
        
        ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search.
        
        ranker: Literal["auto", "default_2024_08_21"]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        results: Optional[List[FileSearchResult]]
        
        The results of the file search.
        
        file_id: str
        
        The ID of the file that result was found in.
        
        file_name: str
        
        The name of the file that result was found in.
        
        score: float
        
        The score of the result. All values must be a floating point number between 0 and 1.
        
        content: Optional[List[FileSearchResultContent]]
        
        The content of the result that was found. The content is only included if requested via the include query parameter.
        
        text: Optional[str]
        
        The text content of the file.
        
        type: Optional[Literal["text"]]
        
        The type of the content.
        
        "text"
        
        type: Literal["file_search"]
        
        The type of tool call. This is always going to be file_search for this type of tool call.
        
        "file_search"
        
        class FunctionToolCall: …
        
        id: str
        
        The ID of the tool call object.
        
        function: Function
        
        The definition of the function that was called.
        
        arguments: str
        
        The arguments passed to the function.
        
        name: str
        
        The name of the function.
        
        output: Optional[str]
        
        The output of the function. This will be null if the outputs have not been submitted yet.
        
        type: Literal["function"]
        
        The type of tool call. This is always going to be function for this type of tool call.
        
        "function"
        
        type: Literal["tool_calls"]
        
        Always tool_calls.
        
        "tool_calls"
    - thread_id: str
      
      The ID of the thread that was run.
    - type: Literal["message_creation", "tool_calls"]
      
      The type of run step, which can be either message_creation or tool_calls.
      - "message_creation"
      - "tool_calls"
    - usage: Optional[Usage]
      
      Usage statistics related to the run step. This value will be null while the run step's status is in_progress.
      - completion_tokens: int
        
        Number of completion tokens used over the course of the run step.
      - prompt_tokens: int
        
        Number of prompt tokens used over the course of the run step.
      - total_tokens: int
        
        Total number of tokens used (prompt + completion).
  - event: Literal["thread.run.step.created"]
    - "thread.run.step.created"
- class ThreadRunStepInProgress: …
  
  Occurs when a run step moves to an in_progress state.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.in_progress"]
    - "thread.run.step.in_progress"
- class ThreadRunStepDelta: …
  
  Occurs when parts of a run step are being streamed.
  - data: RunStepDeltaEvent
    
    Represents a run step delta i.e. any changed fields on a run step during streaming.
    - id: str
      
      The identifier of the run step, which can be referenced in API endpoints.
    - delta: RunStepDelta
      
      The delta containing the fields that have changed on the run step.
      - step_details: Optional[StepDetails]
        
        The details of the run step.
        
        class RunStepDeltaMessageDelta: …
        
        Details of the message creation by the run step.
        
        type: Literal["message_creation"]
        
        Always message_creation.
        
        "message_creation"
        
        message_creation: Optional[MessageCreation]
        
        message_id: Optional[str]
        
        The ID of the message that was created by this run step.
        
        class ToolCallDeltaObject: …
        
        Details of the tool call.
        
        type: Literal["tool_calls"]
        
        Always tool_calls.
        
        "tool_calls"
        
        tool_calls: Optional[List[ToolCallDelta]]
        
        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.
        
        class CodeInterpreterToolCallDelta: …
        
        Details of the Code Interpreter tool call the run step was involved in.
        
        index: int
        
        The index of the tool call in the tool calls array.
        
        type: Literal["code_interpreter"]
        
        The type of tool call. This is always going to be code_interpreter for this type of tool call.
        
        "code_interpreter"
        
        id: Optional[str]
        
        The ID of the tool call.
        
        code_interpreter: Optional[CodeInterpreter]
        
        The Code Interpreter tool call definition.
        
        input: Optional[str]
        
        The input to the Code Interpreter tool call.
        
        outputs: Optional[List[CodeInterpreterOutput]]
        
        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.
        
        class CodeInterpreterLogs: …
        
        Text output from the Code Interpreter tool call as part of a run step.
        
        index: int
        
        The index of the output in the outputs array.
        
        type: Literal["logs"]
        
        Always logs.
        
        "logs"
        
        logs: Optional[str]
        
        The text output from the Code Interpreter tool call.
        
        class CodeInterpreterOutputImage: …
        
        index: int
        
        The index of the output in the outputs array.
        
        type: Literal["image"]
        
        Always image.
        
        "image"
        
        image: Optional[Image]
        
        file_id: Optional[str]
        
        The file ID of the image.
        
        class FileSearchToolCallDelta: …
        
        file_search: object
        
        For now, this is always going to be an empty object.
        
        index: int
        
        The index of the tool call in the tool calls array.
        
        type: Literal["file_search"]
        
        The type of tool call. This is always going to be file_search for this type of tool call.
        
        "file_search"
        
        id: Optional[str]
        
        The ID of the tool call object.
        
        class FunctionToolCallDelta: …
        
        index: int
        
        The index of the tool call in the tool calls array.
        
        type: Literal["function"]
        
        The type of tool call. This is always going to be function for this type of tool call.
        
        "function"
        
        id: Optional[str]
        
        The ID of the tool call object.
        
        function: Optional[Function]
        
        The definition of the function that was called.
        
        arguments: Optional[str]
        
        The arguments passed to the function.
        
        name: Optional[str]
        
        The name of the function.
        
        output: Optional[str]
        
        The output of the function. This will be null if the outputs have not been submitted yet.
    - object: Literal["thread.run.step.delta"]
      
      The object type, which is always thread.run.step.delta.
      - "thread.run.step.delta"
  - event: Literal["thread.run.step.delta"]
    - "thread.run.step.delta"
- class ThreadRunStepCompleted: …
  
  Occurs when a run step is completed.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.completed"]
    - "thread.run.step.completed"
- class ThreadRunStepFailed: …
  
  Occurs when a run step fails.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.failed"]
    - "thread.run.step.failed"
- class ThreadRunStepCancelled: …
  
  Occurs when a run step is cancelled.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.cancelled"]
    - "thread.run.step.cancelled"
- class ThreadRunStepExpired: …
  
  Occurs when a run step expires.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.expired"]
    - "thread.run.step.expired"
- class ThreadMessageCreated: …
  
  Occurs when a message is created.
  - data: Message
    
    Represents a message within a thread.
    - id: str
      
      The identifier, which can be referenced in API endpoints.
    - assistant_id: Optional[str]
      
      If applicable, the ID of the assistant that authored this message.
    - attachments: Optional[List[Attachment]]
      
      A list of files attached to the message, and the tools they were added to.
      - file_id: Optional[str]
        
        The ID of the file to attach to the message.
      - tools: Optional[List[AttachmentTool]]
        
        The tools to add this file to.
        
        class CodeInterpreterTool: …
        
        class AttachmentToolAssistantToolsFileSearchTypeOnly: …
        
        type: Literal["file_search"]
        
        The type of tool being defined: file_search
        
        "file_search"
    - completed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the message was completed.
    - content: List[MessageContent]
      
      The content of the message in array of text and/or images.
      - class ImageFileContentBlock: …
        
        References an image File in the content of a message.
        
        image_file: ImageFile
        
        file_id: str
        
        The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.
        
        "auto"
        
        "low"
        
        "high"
        
        type: Literal["image_file"]
        
        Always image_file.
        
        "image_file"
      - class ImageURLContentBlock: …
        
        References an image URL in the content of a message.
        
        image_url: ImageURL
        
        url: str
        
        The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high. Default value is auto
        
        "auto"
        
        "low"
        
        "high"
        
        type: Literal["image_url"]
        
        The type of the content part.
        
        "image_url"
      - class TextContentBlock: …
        
        The text content that is part of a message.
        
        text: Text
        
        annotations: List[Annotation]
        
        class FileCitationAnnotation: …
        
        A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.
        
        end_index: int
        
        file_citation: FileCitation
        
        file_id: str
        
        The ID of the specific File the citation is from.
        
        start_index: int
        
        text: str
        
        The text in the message content that needs to be replaced.
        
        type: Literal["file_citation"]
        
        Always file_citation.
        
        "file_citation"
        
        class FilePathAnnotation: …
        
        A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.
        
        end_index: int
        
        file_path: FilePath
        
        file_id: str
        
        The ID of the file that was generated.
        
        start_index: int
        
        text: str
        
        The text in the message content that needs to be replaced.
        
        type: Literal["file_path"]
        
        Always file_path.
        
        "file_path"
        
        value: str
        
        The data that makes up the text.
        
        type: Literal["text"]
        
        Always text.
        
        "text"
      - class RefusalContentBlock: …
        
        The refusal content generated by the assistant.
        
        refusal: str
        
        type: Literal["refusal"]
        
        Always refusal.
        
        "refusal"
    - created_at: int
      
      The Unix timestamp (in seconds) for when the message was created.
    - incomplete_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the message was marked as incomplete.
    - incomplete_details: Optional[IncompleteDetails]
      
      On an incomplete message, details about why the message is incomplete.
      - reason: Literal["content_filter", "max_tokens", "run_cancelled", 2 more]
        
        The reason the message is incomplete.
        
        "content_filter"
        
        "max_tokens"
        
        "run_cancelled"
        
        "run_expired"
        
        "run_failed"
    - metadata: Optional[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.
    - object: Literal["thread.message"]
      
      The object type, which is always thread.message.
      - "thread.message"
    - role: Literal["user", "assistant"]
      
      The entity that produced the message. One of user or assistant.
      - "user"
      - "assistant"
    - run_id: Optional[str]
      
      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: Literal["in_progress", "incomplete", "completed"]
      
      The status of the message, which can be either in_progress, incomplete, or completed.
      - "in_progress"
      - "incomplete"
      - "completed"
    - thread_id: str
      
      The thread ID that this message belongs to.
  - event: Literal["thread.message.created"]
    - "thread.message.created"
- class ThreadMessageInProgress: …
  
  Occurs when a message moves to an in_progress state.
  - data: Message
    
    Represents a message within a thread.
  - event: Literal["thread.message.in_progress"]
    - "thread.message.in_progress"
- class ThreadMessageDelta: …
  
  Occurs when parts of a Message are being streamed.
  - data: MessageDeltaEvent
    
    Represents a message delta i.e. any changed fields on a message during streaming.
    - id: str
      
      The identifier of the message, which can be referenced in API endpoints.
    - delta: MessageDelta
      
      The delta containing the fields that have changed on the Message.
      - content: Optional[List[MessageContentDelta]]
        
        The content of the message in array of text and/or images.
        
        class ImageFileDeltaBlock: …
        
        References an image File in the content of a message.
        
        index: int
        
        The index of the content part in the message.
        
        type: Literal["image_file"]
        
        Always image_file.
        
        "image_file"
        
        image_file: Optional[ImageFileDelta]
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.
        
        "auto"
        
        "low"
        
        "high"
        
        file_id: Optional[str]
        
        The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.
        
        class TextDeltaBlock: …
        
        The text content that is part of a message.
        
        index: int
        
        The index of the content part in the message.
        
        type: Literal["text"]
        
        Always text.
        
        "text"
        
        text: Optional[TextDelta]
        
        annotations: Optional[List[AnnotationDelta]]
        
        class FileCitationDeltaAnnotation: …
        
        A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.
        
        index: int
        
        The index of the annotation in the text content part.
        
        type: Literal["file_citation"]
        
        Always file_citation.
        
        "file_citation"
        
        end_index: Optional[int]
        
        file_citation: Optional[FileCitation]
        
        file_id: Optional[str]
        
        The ID of the specific File the citation is from.
        
        quote: Optional[str]
        
        The specific quote in the file.
        
        start_index: Optional[int]
        
        text: Optional[str]
        
        The text in the message content that needs to be replaced.
        
        class FilePathDeltaAnnotation: …
        
        A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.
        
        index: int
        
        The index of the annotation in the text content part.
        
        type: Literal["file_path"]
        
        Always file_path.
        
        "file_path"
        
        end_index: Optional[int]
        
        file_path: Optional[FilePath]
        
        file_id: Optional[str]
        
        The ID of the file that was generated.
        
        start_index: Optional[int]
        
        text: Optional[str]
        
        The text in the message content that needs to be replaced.
        
        value: Optional[str]
        
        The data that makes up the text.
        
        class RefusalDeltaBlock: …
        
        The refusal content that is part of a message.
        
        index: int
        
        The index of the refusal part in the message.
        
        type: Literal["refusal"]
        
        Always refusal.
        
        "refusal"
        
        refusal: Optional[str]
        
        class ImageURLDeltaBlock: …
        
        References an image URL in the content of a message.
        
        index: int
        
        The index of the content part in the message.
        
        type: Literal["image_url"]
        
        Always image_url.
        
        "image_url"
        
        image_url: Optional[ImageURLDelta]
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high.
        
        "auto"
        
        "low"
        
        "high"
        
        url: Optional[str]
        
        The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.
      - role: Optional[Literal["user", "assistant"]]
        
        The entity that produced the message. One of user or assistant.
        
        "user"
        
        "assistant"
    - object: Literal["thread.message.delta"]
      
      The object type, which is always thread.message.delta.
      - "thread.message.delta"
  - event: Literal["thread.message.delta"]
    - "thread.message.delta"
- class ThreadMessageCompleted: …
  
  Occurs when a message is completed.
  - data: Message
    
    Represents a message within a thread.
  - event: Literal["thread.message.completed"]
    - "thread.message.completed"
- class ThreadMessageIncomplete: …
  
  Occurs when a message ends before it is completed.
  - data: Message
    
    Represents a message within a thread.
  - event: Literal["thread.message.incomplete"]
    - "thread.message.incomplete"
- class ErrorEvent: …
  
  Occurs when an error occurs. This can happen due to an internal server error or a timeout.
  - data: ErrorObject
    - code: Optional[str]
    - message: str
    - param: Optional[str]
    - type: str
  - event: Literal["error"]
    - "error"

Assistant Tool

AssistantTool
- class CodeInterpreterTool: …
  - type: Literal["code_interpreter"]
    
    The type of tool being defined: code_interpreter
    - "code_interpreter"
- class FileSearchTool: …
  - type: Literal["file_search"]
    
    The type of tool being defined: file_search
    - "file_search"
  - file_search: Optional[FileSearch]
    
    Overrides for the file search tool.
    - max_num_results: Optional[int]
      
      The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
      
      Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
    - ranking_options: Optional[FileSearchRankingOptions]
      
      The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
      
      See the file search tool documentation for more information.
      - score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
      - ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
- class FunctionTool: …
  - function: FunctionDefinition
    - name: str
      
      The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
    - description: Optional[str]
      
      A description of what the function does, used by the model to choose when and how to call the function.
    - parameters: Optional[FunctionParameters]
      
      The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
      
      Omitting parameters defines a function with an empty parameter list.
    - strict: Optional[bool]
      
      Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
  - type: Literal["function"]
    
    The type of tool being defined: function
    - "function"

Code Interpreter Tool

class CodeInterpreterTool: …
- type: Literal["code_interpreter"]
  
  The type of tool being defined: code_interpreter
  - "code_interpreter"

File Search Tool

class FileSearchTool: …
- type: Literal["file_search"]
  
  The type of tool being defined: file_search
  - "file_search"
- file_search: Optional[FileSearch]
  
  Overrides for the file search tool.
  - max_num_results: Optional[int]
    
    The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
    
    Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
  - ranking_options: Optional[FileSearchRankingOptions]
    
    The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
    
    See the file search tool documentation for more information.
    - score_threshold: float
      
      The score threshold for the file search. All values must be a floating point number between 0 and 1.
    - ranker: Optional[Literal["auto", "default_2024_08_21"]]
      
      The ranker to use for the file search. If not specified will use the auto ranker.
      - "auto"
      - "default_2024_08_21"

Function Tool

class FunctionTool: …
- function: FunctionDefinition
  - name: str
    
    The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
  - description: Optional[str]
    
    A description of what the function does, used by the model to choose when and how to call the function.
  - parameters: Optional[FunctionParameters]
    
    The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
    
    Omitting parameters defines a function with an empty parameter list.
  - strict: Optional[bool]
    
    Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
- type: Literal["function"]
  
  The type of tool being defined: function
  - "function"

Message Stream Event

MessageStreamEvent

Occurs when a message is created.
- class ThreadMessageCreated: …
  
  Occurs when a message is created.
  - data: Message
    
    Represents a message within a thread.
    - id: str
      
      The identifier, which can be referenced in API endpoints.
    - assistant_id: Optional[str]
      
      If applicable, the ID of the assistant that authored this message.
    - attachments: Optional[List[Attachment]]
      
      A list of files attached to the message, and the tools they were added to.
      - file_id: Optional[str]
        
        The ID of the file to attach to the message.
      - tools: Optional[List[AttachmentTool]]
        
        The tools to add this file to.
        
        class CodeInterpreterTool: …
        
        type: Literal["code_interpreter"]
        
        The type of tool being defined: code_interpreter
        
        "code_interpreter"
        
        class AttachmentToolAssistantToolsFileSearchTypeOnly: …
        
        type: Literal["file_search"]
        
        The type of tool being defined: file_search
        
        "file_search"
    - completed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the message was completed.
    - content: List[MessageContent]
      
      The content of the message in array of text and/or images.
      - class ImageFileContentBlock: …
        
        References an image File in the content of a message.
        
        image_file: ImageFile
        
        file_id: str
        
        The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.
        
        "auto"
        
        "low"
        
        "high"
        
        type: Literal["image_file"]
        
        Always image_file.
        
        "image_file"
      - class ImageURLContentBlock: …
        
        References an image URL in the content of a message.
        
        image_url: ImageURL
        
        url: str
        
        The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high. Default value is auto
        
        "auto"
        
        "low"
        
        "high"
        
        type: Literal["image_url"]
        
        The type of the content part.
        
        "image_url"
      - class TextContentBlock: …
        
        The text content that is part of a message.
        
        text: Text
        
        annotations: List[Annotation]
        
        class FileCitationAnnotation: …
        
        A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.
        
        end_index: int
        
        file_citation: FileCitation
        
        file_id: str
        
        The ID of the specific File the citation is from.
        
        start_index: int
        
        text: str
        
        The text in the message content that needs to be replaced.
        
        type: Literal["file_citation"]
        
        Always file_citation.
        
        "file_citation"
        
        class FilePathAnnotation: …
        
        A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.
        
        end_index: int
        
        file_path: FilePath
        
        file_id: str
        
        The ID of the file that was generated.
        
        start_index: int
        
        text: str
        
        The text in the message content that needs to be replaced.
        
        type: Literal["file_path"]
        
        Always file_path.
        
        "file_path"
        
        value: str
        
        The data that makes up the text.
        
        type: Literal["text"]
        
        Always text.
        
        "text"
      - class RefusalContentBlock: …
        
        The refusal content generated by the assistant.
        
        refusal: str
        
        type: Literal["refusal"]
        
        Always refusal.
        
        "refusal"
    - created_at: int
      
      The Unix timestamp (in seconds) for when the message was created.
    - incomplete_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the message was marked as incomplete.
    - incomplete_details: Optional[IncompleteDetails]
      
      On an incomplete message, details about why the message is incomplete.
      - reason: Literal["content_filter", "max_tokens", "run_cancelled", 2 more]
        
        The reason the message is incomplete.
        
        "content_filter"
        
        "max_tokens"
        
        "run_cancelled"
        
        "run_expired"
        
        "run_failed"
    - metadata: Optional[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.
    - object: Literal["thread.message"]
      
      The object type, which is always thread.message.
      - "thread.message"
    - role: Literal["user", "assistant"]
      
      The entity that produced the message. One of user or assistant.
      - "user"
      - "assistant"
    - run_id: Optional[str]
      
      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: Literal["in_progress", "incomplete", "completed"]
      
      The status of the message, which can be either in_progress, incomplete, or completed.
      - "in_progress"
      - "incomplete"
      - "completed"
    - thread_id: str
      
      The thread ID that this message belongs to.
  - event: Literal["thread.message.created"]
    - "thread.message.created"
- class ThreadMessageInProgress: …
  
  Occurs when a message moves to an in_progress state.
  - data: Message
    
    Represents a message within a thread.
  - event: Literal["thread.message.in_progress"]
    - "thread.message.in_progress"
- class ThreadMessageDelta: …
  
  Occurs when parts of a Message are being streamed.
  - data: MessageDeltaEvent
    
    Represents a message delta i.e. any changed fields on a message during streaming.
    - id: str
      
      The identifier of the message, which can be referenced in API endpoints.
    - delta: MessageDelta
      
      The delta containing the fields that have changed on the Message.
      - content: Optional[List[MessageContentDelta]]
        
        The content of the message in array of text and/or images.
        
        class ImageFileDeltaBlock: …
        
        References an image File in the content of a message.
        
        index: int
        
        The index of the content part in the message.
        
        type: Literal["image_file"]
        
        Always image_file.
        
        "image_file"
        
        image_file: Optional[ImageFileDelta]
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image if specified by the user. low uses fewer tokens, you can opt in to high resolution using high.
        
        "auto"
        
        "low"
        
        "high"
        
        file_id: Optional[str]
        
        The File ID of the image in the message content. Set purpose="vision" when uploading the File if you need to later display the file content.
        
        class TextDeltaBlock: …
        
        The text content that is part of a message.
        
        index: int
        
        The index of the content part in the message.
        
        type: Literal["text"]
        
        Always text.
        
        "text"
        
        text: Optional[TextDelta]
        
        annotations: Optional[List[AnnotationDelta]]
        
        class FileCitationDeltaAnnotation: …
        
        A citation within the message that points to a specific quote from a specific File associated with the assistant or the message. Generated when the assistant uses the "file_search" tool to search files.
        
        index: int
        
        The index of the annotation in the text content part.
        
        type: Literal["file_citation"]
        
        Always file_citation.
        
        "file_citation"
        
        end_index: Optional[int]
        
        file_citation: Optional[FileCitation]
        
        file_id: Optional[str]
        
        The ID of the specific File the citation is from.
        
        quote: Optional[str]
        
        The specific quote in the file.
        
        start_index: Optional[int]
        
        text: Optional[str]
        
        The text in the message content that needs to be replaced.
        
        class FilePathDeltaAnnotation: …
        
        A URL for the file that's generated when the assistant used the code_interpreter tool to generate a file.
        
        index: int
        
        The index of the annotation in the text content part.
        
        type: Literal["file_path"]
        
        Always file_path.
        
        "file_path"
        
        end_index: Optional[int]
        
        file_path: Optional[FilePath]
        
        file_id: Optional[str]
        
        The ID of the file that was generated.
        
        start_index: Optional[int]
        
        text: Optional[str]
        
        The text in the message content that needs to be replaced.
        
        value: Optional[str]
        
        The data that makes up the text.
        
        class RefusalDeltaBlock: …
        
        The refusal content that is part of a message.
        
        index: int
        
        The index of the refusal part in the message.
        
        type: Literal["refusal"]
        
        Always refusal.
        
        "refusal"
        
        refusal: Optional[str]
        
        class ImageURLDeltaBlock: …
        
        References an image URL in the content of a message.
        
        index: int
        
        The index of the content part in the message.
        
        type: Literal["image_url"]
        
        Always image_url.
        
        "image_url"
        
        image_url: Optional[ImageURLDelta]
        
        detail: Optional[Literal["auto", "low", "high"]]
        
        Specifies the detail level of the image. low uses fewer tokens, you can opt in to high resolution using high.
        
        "auto"
        
        "low"
        
        "high"
        
        url: Optional[str]
        
        The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.
      - role: Optional[Literal["user", "assistant"]]
        
        The entity that produced the message. One of user or assistant.
        
        "user"
        
        "assistant"
    - object: Literal["thread.message.delta"]
      
      The object type, which is always thread.message.delta.
      - "thread.message.delta"
  - event: Literal["thread.message.delta"]
    - "thread.message.delta"
- class ThreadMessageCompleted: …
  
  Occurs when a message is completed.
  - data: Message
    
    Represents a message within a thread.
  - event: Literal["thread.message.completed"]
    - "thread.message.completed"
- class ThreadMessageIncomplete: …
  
  Occurs when a message ends before it is completed.
  - data: Message
    
    Represents a message within a thread.
  - event: Literal["thread.message.incomplete"]
    - "thread.message.incomplete"

Run Step Stream Event

RunStepStreamEvent

Occurs when a run step is created.
- class ThreadRunStepCreated: …
  
  Occurs when a run step is created.
  - data: RunStep
    
    Represents a step in execution of a run.
    - id: str
      
      The identifier of the run step, which can be referenced in API endpoints.
    - assistant_id: str
      
      The ID of the assistant associated with the run step.
    - cancelled_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step was cancelled.
    - completed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step completed.
    - created_at: int
      
      The Unix timestamp (in seconds) for when the run step was created.
    - expired_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.
    - failed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run step failed.
    - last_error: Optional[LastError]
      
      The last error associated with this run step. Will be null if there are no errors.
      - code: Literal["server_error", "rate_limit_exceeded"]
        
        One of server_error or rate_limit_exceeded.
        
        "server_error"
        
        "rate_limit_exceeded"
      - message: str
        
        A human-readable description of the error.
    - metadata: Optional[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.
    - object: Literal["thread.run.step"]
      
      The object type, which is always thread.run.step.
      - "thread.run.step"
    - run_id: str
      
      The ID of the run that this run step is a part of.
    - status: Literal["in_progress", "cancelled", "failed", 2 more]
      
      The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.
      - "in_progress"
      - "cancelled"
      - "failed"
      - "completed"
      - "expired"
    - step_details: StepDetails
      
      The details of the run step.
      - class MessageCreationStepDetails: …
        
        Details of the message creation by the run step.
        
        message_creation: MessageCreation
        
        message_id: str
        
        The ID of the message that was created by this run step.
        
        type: Literal["message_creation"]
        
        Always message_creation.
        
        "message_creation"
      - class ToolCallsStepDetails: …
        
        Details of the tool call.
        
        tool_calls: List[ToolCall]
        
        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.
        
        class CodeInterpreterToolCall: …
        
        Details of the Code Interpreter tool call the run step was involved in.
        
        id: str
        
        The ID of the tool call.
        
        code_interpreter: CodeInterpreter
        
        The Code Interpreter tool call definition.
        
        input: str
        
        The input to the Code Interpreter tool call.
        
        outputs: List[CodeInterpreterOutput]
        
        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.
        
        class CodeInterpreterOutputLogs: …
        
        Text output from the Code Interpreter tool call as part of a run step.
        
        logs: str
        
        The text output from the Code Interpreter tool call.
        
        type: Literal["logs"]
        
        Always logs.
        
        "logs"
        
        class CodeInterpreterOutputImage: …
        
        image: CodeInterpreterOutputImageImage
        
        file_id: str
        
        The file ID of the image.
        
        type: Literal["image"]
        
        Always image.
        
        "image"
        
        type: Literal["code_interpreter"]
        
        The type of tool call. This is always going to be code_interpreter for this type of tool call.
        
        "code_interpreter"
        
        class FileSearchToolCall: …
        
        id: str
        
        The ID of the tool call object.
        
        file_search: FileSearch
        
        For now, this is always going to be an empty object.
        
        ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search.
        
        ranker: Literal["auto", "default_2024_08_21"]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        results: Optional[List[FileSearchResult]]
        
        The results of the file search.
        
        file_id: str
        
        The ID of the file that result was found in.
        
        file_name: str
        
        The name of the file that result was found in.
        
        score: float
        
        The score of the result. All values must be a floating point number between 0 and 1.
        
        content: Optional[List[FileSearchResultContent]]
        
        The content of the result that was found. The content is only included if requested via the include query parameter.
        
        text: Optional[str]
        
        The text content of the file.
        
        type: Optional[Literal["text"]]
        
        The type of the content.
        
        "text"
        
        type: Literal["file_search"]
        
        The type of tool call. This is always going to be file_search for this type of tool call.
        
        "file_search"
        
        class FunctionToolCall: …
        
        id: str
        
        The ID of the tool call object.
        
        function: Function
        
        The definition of the function that was called.
        
        arguments: str
        
        The arguments passed to the function.
        
        name: str
        
        The name of the function.
        
        output: Optional[str]
        
        The output of the function. This will be null if the outputs have not been submitted yet.
        
        type: Literal["function"]
        
        The type of tool call. This is always going to be function for this type of tool call.
        
        "function"
        
        type: Literal["tool_calls"]
        
        Always tool_calls.
        
        "tool_calls"
    - thread_id: str
      
      The ID of the thread that was run.
    - type: Literal["message_creation", "tool_calls"]
      
      The type of run step, which can be either message_creation or tool_calls.
      - "message_creation"
      - "tool_calls"
    - usage: Optional[Usage]
      
      Usage statistics related to the run step. This value will be null while the run step's status is in_progress.
      - completion_tokens: int
        
        Number of completion tokens used over the course of the run step.
      - prompt_tokens: int
        
        Number of prompt tokens used over the course of the run step.
      - total_tokens: int
        
        Total number of tokens used (prompt + completion).
  - event: Literal["thread.run.step.created"]
    - "thread.run.step.created"
- class ThreadRunStepInProgress: …
  
  Occurs when a run step moves to an in_progress state.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.in_progress"]
    - "thread.run.step.in_progress"
- class ThreadRunStepDelta: …
  
  Occurs when parts of a run step are being streamed.
  - data: RunStepDeltaEvent
    
    Represents a run step delta i.e. any changed fields on a run step during streaming.
    - id: str
      
      The identifier of the run step, which can be referenced in API endpoints.
    - delta: RunStepDelta
      
      The delta containing the fields that have changed on the run step.
      - step_details: Optional[StepDetails]
        
        The details of the run step.
        
        class RunStepDeltaMessageDelta: …
        
        Details of the message creation by the run step.
        
        type: Literal["message_creation"]
        
        Always message_creation.
        
        "message_creation"
        
        message_creation: Optional[MessageCreation]
        
        message_id: Optional[str]
        
        The ID of the message that was created by this run step.
        
        class ToolCallDeltaObject: …
        
        Details of the tool call.
        
        type: Literal["tool_calls"]
        
        Always tool_calls.
        
        "tool_calls"
        
        tool_calls: Optional[List[ToolCallDelta]]
        
        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.
        
        class CodeInterpreterToolCallDelta: …
        
        Details of the Code Interpreter tool call the run step was involved in.
        
        index: int
        
        The index of the tool call in the tool calls array.
        
        type: Literal["code_interpreter"]
        
        The type of tool call. This is always going to be code_interpreter for this type of tool call.
        
        "code_interpreter"
        
        id: Optional[str]
        
        The ID of the tool call.
        
        code_interpreter: Optional[CodeInterpreter]
        
        The Code Interpreter tool call definition.
        
        input: Optional[str]
        
        The input to the Code Interpreter tool call.
        
        outputs: Optional[List[CodeInterpreterOutput]]
        
        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.
        
        class CodeInterpreterLogs: …
        
        Text output from the Code Interpreter tool call as part of a run step.
        
        index: int
        
        The index of the output in the outputs array.
        
        type: Literal["logs"]
        
        Always logs.
        
        "logs"
        
        logs: Optional[str]
        
        The text output from the Code Interpreter tool call.
        
        class CodeInterpreterOutputImage: …
        
        index: int
        
        The index of the output in the outputs array.
        
        type: Literal["image"]
        
        Always image.
        
        "image"
        
        image: Optional[Image]
        
        file_id: Optional[str]
        
        The file ID of the image.
        
        class FileSearchToolCallDelta: …
        
        file_search: object
        
        For now, this is always going to be an empty object.
        
        index: int
        
        The index of the tool call in the tool calls array.
        
        type: Literal["file_search"]
        
        The type of tool call. This is always going to be file_search for this type of tool call.
        
        "file_search"
        
        id: Optional[str]
        
        The ID of the tool call object.
        
        class FunctionToolCallDelta: …
        
        index: int
        
        The index of the tool call in the tool calls array.
        
        type: Literal["function"]
        
        The type of tool call. This is always going to be function for this type of tool call.
        
        "function"
        
        id: Optional[str]
        
        The ID of the tool call object.
        
        function: Optional[Function]
        
        The definition of the function that was called.
        
        arguments: Optional[str]
        
        The arguments passed to the function.
        
        name: Optional[str]
        
        The name of the function.
        
        output: Optional[str]
        
        The output of the function. This will be null if the outputs have not been submitted yet.
    - object: Literal["thread.run.step.delta"]
      
      The object type, which is always thread.run.step.delta.
      - "thread.run.step.delta"
  - event: Literal["thread.run.step.delta"]
    - "thread.run.step.delta"
- class ThreadRunStepCompleted: …
  
  Occurs when a run step is completed.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.completed"]
    - "thread.run.step.completed"
- class ThreadRunStepFailed: …
  
  Occurs when a run step fails.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.failed"]
    - "thread.run.step.failed"
- class ThreadRunStepCancelled: …
  
  Occurs when a run step is cancelled.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.cancelled"]
    - "thread.run.step.cancelled"
- class ThreadRunStepExpired: …
  
  Occurs when a run step expires.
  - data: RunStep
    
    Represents a step in execution of a run.
  - event: Literal["thread.run.step.expired"]
    - "thread.run.step.expired"

Run Stream Event

RunStreamEvent

Occurs when a new run is created.
- class ThreadRunCreated: …
  
  Occurs when a new run is created.
  - data: Run
    
    Represents an execution run on a thread.
    - id: str
      
      The identifier, which can be referenced in API endpoints.
    - assistant_id: str
      
      The ID of the assistant used for execution of this run.
    - cancelled_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run was cancelled.
    - completed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run was completed.
    - created_at: int
      
      The Unix timestamp (in seconds) for when the run was created.
    - expires_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run will expire.
    - failed_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run failed.
    - incomplete_details: Optional[IncompleteDetails]
      
      Details on why the run is incomplete. Will be null if the run is not incomplete.
      - reason: Optional[Literal["max_completion_tokens", "max_prompt_tokens"]]
        
        The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.
        
        "max_completion_tokens"
        
        "max_prompt_tokens"
    - instructions: str
      
      The instructions that the assistant used for this run.
    - last_error: Optional[LastError]
      
      The last error associated with this run. Will be null if there are no errors.
      - code: Literal["server_error", "rate_limit_exceeded", "invalid_prompt"]
        
        One of server_error, rate_limit_exceeded, or invalid_prompt.
        
        "server_error"
        
        "rate_limit_exceeded"
        
        "invalid_prompt"
      - message: str
        
        A human-readable description of the error.
    - max_completion_tokens: Optional[int]
      
      The maximum number of completion tokens specified to have been used over the course of the run.
    - max_prompt_tokens: Optional[int]
      
      The maximum number of prompt tokens specified to have been used over the course of the run.
    - metadata: Optional[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.
    - model: str
      
      The model that the assistant used for this run.
    - object: Literal["thread.run"]
      
      The object type, which is always thread.run.
      - "thread.run"
    - parallel_tool_calls: bool
      
      Whether to enable parallel function calling during tool use.
    - required_action: Optional[RequiredAction]
      
      Details on the action required to continue the run. Will be null if no action is required.
      - submit_tool_outputs: RequiredActionSubmitToolOutputs
        
        Details on the tool outputs needed for this run to continue.
        
        tool_calls: List[RequiredActionFunctionToolCall]
        
        A list of the relevant tool calls.
        
        id: str
        
        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.
        
        arguments: str
        
        The arguments that the model expects you to pass to the function.
        
        name: str
        
        The name of the function.
        
        type: Literal["function"]
        
        The type of tool call the output is required for. For now, this is always function.
        
        "function"
      - type: Literal["submit_tool_outputs"]
        
        For now, this is always submit_tool_outputs.
        
        "submit_tool_outputs"
    - response_format: Optional[AssistantResponseFormatOption]
      
      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.
      - Literal["auto"]
        
        auto is the default value
        
        "auto"
      - class ResponseFormatText: …
        
        Default response format. Used to generate text responses.
        
        type: Literal["text"]
        
        The type of response format being defined. Always 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.
        
        type: Literal["json_object"]
        
        The type of response format being defined. Always json_object.
        
        "json_object"
      - class ResponseFormatJSONSchema: …
        
        JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
        
        json_schema: JSONSchema
        
        Structured Outputs configuration options, including a JSON Schema.
        
        name: str
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
        
        description: Optional[str]
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
        
        schema: Optional[Dict[str, object]]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
        
        strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the output. If set to true, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is true. To learn more, read the Structured Outputs guide.
        
        type: Literal["json_schema"]
        
        The type of response format being defined. Always json_schema.
        
        "json_schema"
    - started_at: Optional[int]
      
      The Unix timestamp (in seconds) for when the run was started.
    - status: RunStatus
      
      The status of the run, which can be either queued, in_progress, requires_action, cancelling, cancelled, failed, completed, incomplete, or expired.
      - "queued"
      - "in_progress"
      - "requires_action"
      - "cancelling"
      - "cancelled"
      - "failed"
      - "completed"
      - "incomplete"
      - "expired"
    - thread_id: str
      
      The ID of the thread that was executed on as a part of this run.
    - tool_choice: Optional[AssistantToolChoiceOption]
      
      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.
      - Literal["none", "auto", "required"]
        
        none means the model will not call any tools and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools before responding to the user.
        
        "none"
        
        "auto"
        
        "required"
      - class AssistantToolChoice: …
        
        Specifies a tool the model should use. Use to force the model to call a specific tool.
        
        type: Literal["function", "code_interpreter", "file_search"]
        
        The type of the tool. If type is function, the function name must be set
        
        "function"
        
        "code_interpreter"
        
        "file_search"
        
        function: Optional[AssistantToolChoiceFunction]
        
        name: str
        
        The name of the function to call.
    - tools: List[AssistantTool]
      
      The list of tools that the assistant used for this run.
      - class CodeInterpreterTool: …
        
        type: Literal["code_interpreter"]
        
        The type of tool being defined: code_interpreter
        
        "code_interpreter"
      - class FileSearchTool: …
        
        type: Literal["file_search"]
        
        The type of tool being defined: file_search
        
        "file_search"
        
        file_search: Optional[FileSearch]
        
        Overrides for the file search tool.
        
        max_num_results: Optional[int]
        
        The maximum number of results the file search tool should output. The default is 20 for gpt-4* models and 5 for gpt-3.5-turbo. This number should be between 1 and 50 inclusive.
        
        Note that the file search tool may output fewer than max_num_results results. See the file search tool documentation for more information.
        
        ranking_options: Optional[FileSearchRankingOptions]
        
        The ranking options for the file search. If not specified, the file search tool will use the auto ranker and a score_threshold of 0.
        
        See the file search tool documentation for more information.
        
        score_threshold: float
        
        The score threshold for the file search. All values must be a floating point number between 0 and 1.
        
        ranker: Optional[Literal["auto", "default_2024_08_21"]]
        
        The ranker to use for the file search. If not specified will use the auto ranker.
        
        "auto"
        
        "default_2024_08_21"
      - class FunctionTool: …
        
        function: FunctionDefinition
        
        name: str
        
        The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
        
        description: Optional[str]
        
        A description of what the function does, used by the model to choose when and how to call the function.
        
        parameters: Optional[FunctionParameters]
        
        The parameters the functions accepts, described as a JSON Schema object. See the guide for examples, and the JSON Schema reference for documentation about the format.
        
        Omitting parameters defines a function with an empty parameter list.
        
        strict: Optional[bool]
        
        Whether to enable strict schema adherence when generating the function call. If set to true, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is true. Learn more about Structured Outputs in the function calling guide.
        
        type: Literal["function"]
        
        The type of tool being defined: function
        
        "function"
    - truncation_strategy: Optional[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: Literal["auto", "last_messages"]
        
        The truncation strategy to use for the thread. The default is auto. If set to last_messages, the thread will be truncated to the n most recent messages in the thread. When set to auto, messages in the middle of the thread will be dropped to fit the context length of the model, max_prompt_tokens.
        
        "auto"
        
        "last_messages"
      - last_messages: Optional[int]
        
        The number of most recent messages from the thread when constructing the context for the run.
    - usage: Optional[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.).
      - completion_tokens: int
        
        Number of completion tokens used over the course of the run.
      - prompt_tokens: int
        
        Number of prompt tokens used over the course of the run.
      - total_tokens: int
        
        Total number of tokens used (prompt + completion).
    - temperature: Optional[float]
      
      The sampling temperature used for this run. If not set, defaults to 1.
    - top_p: Optional[float]
      
      The nucleus sampling value used for this run. If not set, defaults to 1.
  - event: Literal["thread.run.created"]
    - "thread.run.created"
- class ThreadRunQueued: …
  
  Occurs when a run moves to a queued status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.queued"]
    - "thread.run.queued"
- class ThreadRunInProgress: …
  
  Occurs when a run moves to an in_progress status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.in_progress"]
    - "thread.run.in_progress"
- class ThreadRunRequiresAction: …
  
  Occurs when a run moves to a requires_action status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.requires_action"]
    - "thread.run.requires_action"
- class ThreadRunCompleted: …
  
  Occurs when a run is completed.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.completed"]
    - "thread.run.completed"
- class ThreadRunIncomplete: …
  
  Occurs when a run ends with status incomplete.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.incomplete"]
    - "thread.run.incomplete"
- class ThreadRunFailed: …
  
  Occurs when a run fails.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.failed"]
    - "thread.run.failed"
- class ThreadRunCancelling: …
  
  Occurs when a run moves to a cancelling status.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.cancelling"]
    - "thread.run.cancelling"
- class ThreadRunCancelled: …
  
  Occurs when a run is cancelled.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.cancelled"]
    - "thread.run.cancelled"
- class ThreadRunExpired: …
  
  Occurs when a run expires.
  - data: Run
    
    Represents an execution run on a thread.
  - event: Literal["thread.run.expired"]
    - "thread.run.expired"

Thread Stream Event

class ThreadStreamEvent: …

Occurs when a new thread is created.
- data: Thread
  
  Represents a thread that contains messages.
  - id: str
    
    The identifier, which can be referenced in API endpoints.
  - created_at: int
    
    The Unix timestamp (in seconds) for when the thread was created.
  - metadata: Optional[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.
  - object: Literal["thread"]
    
    The object type, which is always thread.
    - "thread"
  - tool_resources: Optional[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.
    - code_interpreter: Optional[ToolResourcesCodeInterpreter]
      - file_ids: Optional[List[str]]
        
        A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
    - file_search: Optional[ToolResourcesFileSearch]
      - vector_store_ids: Optional[List[str]]
        
        The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.
- event: Literal["thread.created"]
  - "thread.created"
- enabled: Optional[bool]
  
  Whether to enable input audio transcription.

python/resources/beta/subresources/assistants/index.md +2365 −431

6 6

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

8 8

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

10 10

11### Parameters11### Parameters

12 12

75 75

76 - `"assistant"`76 - `"assistant"`

77 77

~~78 - `tools: List[object]`~~78 - `tools: List[AssistantTool]`

79 79

80 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`.80 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`.

81 81

82 - `class CodeInterpreterTool: …`

84 - `type: Literal["code_interpreter"]`

86 The type of tool being defined: `code_interpreter`

88 - `"code_interpreter"`

90 - `class FileSearchTool: …`

92 - `type: Literal["file_search"]`

94 The type of tool being defined: `file_search`

96 - `"file_search"`

98 - `file_search: Optional[FileSearch]`

100 Overrides for the file search tool.

101

102 - `max_num_results: Optional[int]`

103

104 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.

105

106 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.

107

108 - `ranking_options: Optional[FileSearchRankingOptions]`

109

110 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.

111

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

113

114 - `score_threshold: float`

115

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

117

118 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

119

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

121

122 - `"auto"`

123

124 - `"default_2024_08_21"`

125

126 - `class FunctionTool: …`

127

128 - `function: FunctionDefinition`

129

130 - `name: str`

131

132 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.

133

134 - `description: Optional[str]`

135

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

137

138 - `parameters: Optional[FunctionParameters]`

139

140 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.

141

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

143

144 - `strict: Optional[bool]`

145

146 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).

147

148 - `type: Literal["function"]`

149

150 The type of tool being defined: `function`

151

152 - `"function"`

153

82 - `response_format: Optional[AssistantResponseFormatOption]`154 - `response_format: Optional[AssistantResponseFormatOption]`

83 155

84 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`.156 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`.

213 "name": "name",285 "name": "name",

214 "object": "assistant",286 "object": "assistant",

215 "tools": [287 "tools": [

216 {}288 {

289 "type": "code_interpreter"

290 }

217 ],291 ],

218 "response_format": "auto",292 "response_format": "auto",

219 "temperature": 1,293 "temperature": 1,

316 390

317**post** `/assistants`391**post** `/assistants`

318 392

319Create assistant393Create an assistant with a model and instructions.

320 394

321### Parameters395### Parameters

322 396

683 Keys are strings with a maximum length of 64 characters. Values are strings757 Keys are strings with a maximum length of 64 characters. Values are strings

684 with a maximum length of 512 characters.758 with a maximum length of 512 characters.

685 759

686- `tools: Optional[Iterable[object]]`760- `tools: Optional[Iterable[AssistantToolParam]]`

687 761

688 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`.762 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`.

689 763

764 - `class CodeInterpreterTool: …`

765

766 - `type: Literal["code_interpreter"]`

767

768 The type of tool being defined: `code_interpreter`

769

770 - `"code_interpreter"`

771

772 - `class FileSearchTool: …`

773

774 - `type: Literal["file_search"]`

775

776 The type of tool being defined: `file_search`

777

778 - `"file_search"`

779

780 - `file_search: Optional[FileSearch]`

781

782 Overrides for the file search tool.

783

784 - `max_num_results: Optional[int]`

785

786 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.

787

788 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.

789

790 - `ranking_options: Optional[FileSearchRankingOptions]`

791

792 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.

793

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

795

796 - `score_threshold: float`

797

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

799

800 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

801

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

803

804 - `"auto"`

805

806 - `"default_2024_08_21"`

807

808 - `class FunctionTool: …`

809

810 - `function: FunctionDefinition`

811

812 - `name: str`

813

814 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.

815

816 - `description: Optional[str]`

817

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

819

820 - `parameters: Optional[FunctionParameters]`

821

822 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.

823

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

825

826 - `strict: Optional[bool]`

827

828 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).

829

830 - `type: Literal["function"]`

831

832 The type of tool being defined: `function`

833

834 - `"function"`

835

690- `top_p: Optional[float]`836- `top_p: Optional[float]`

691 837

692 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.838 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.

738 884

739 - `"assistant"`885 - `"assistant"`

740 886

741 - `tools: List[object]`887 - `tools: List[AssistantTool]`

742 888

743 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`.889 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`.

744 890

891 - `class CodeInterpreterTool: …`

892

893 - `type: Literal["code_interpreter"]`

894

895 The type of tool being defined: `code_interpreter`

896

897 - `"code_interpreter"`

898

899 - `class FileSearchTool: …`

900

901 - `type: Literal["file_search"]`

902

903 The type of tool being defined: `file_search`

904

905 - `"file_search"`

906

907 - `file_search: Optional[FileSearch]`

908

909 Overrides for the file search tool.

910

911 - `max_num_results: Optional[int]`

912

913 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.

914

915 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.

916

917 - `ranking_options: Optional[FileSearchRankingOptions]`

918

919 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.

920

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

922

923 - `score_threshold: float`

924

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

926

927 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

928

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

930

931 - `"auto"`

932

933 - `"default_2024_08_21"`

934

935 - `class FunctionTool: …`

936

937 - `function: FunctionDefinition`

938

939 - `name: str`

940

941 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.

942

943 - `description: Optional[str]`

944

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

946

947 - `parameters: Optional[FunctionParameters]`

948

949 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.

950

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

952

953 - `strict: Optional[bool]`

954

955 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).

956

957 - `type: Literal["function"]`

958

959 The type of tool being defined: `function`

960

961 - `"function"`

962

745 - `response_format: Optional[AssistantResponseFormatOption]`963 - `response_format: Optional[AssistantResponseFormatOption]`

746 964

747 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`.965 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`.

875 "name": "name",1093 "name": "name",

876 "object": "assistant",1094 "object": "assistant",

877 "tools": [1095 "tools": [

878 {}1096 {

1097 "type": "code_interpreter"

1098 }

879 ],1099 ],

880 "response_format": "auto",1100 "response_format": "auto",

881 "temperature": 1,1101 "temperature": 1,

983 1203

984**get** `/assistants/{assistant_id}`1204**get** `/assistants/{assistant_id}`

985 1205

986Retrieve assistant1206Retrieves an assistant.

987 1207

988### Parameters1208### Parameters

989 1209

1034 1254

1035 - `"assistant"`1255 - `"assistant"`

1036 1256

1037 - `tools: List[object]`1257 - `tools: List[AssistantTool]`

1038 1258

1039 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`.1259 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`.

1040 1260

1041 - `response_format: Optional[AssistantResponseFormatOption]`1261 - `class CodeInterpreterTool: …`

1042 1262

1043 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`.1263 - `type: Literal["code_interpreter"]`

1044 1264

1045 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).1265 The type of tool being defined: `code_interpreter`

1046 1266

1047 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.1267 - `"code_interpreter"`

1048 1268

1049 **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.1269 - `class FileSearchTool: …`

1050 1270

1051 - `Literal["auto"]`1271 - `type: Literal["file_search"]`

1052 1272

1053 `auto` is the default value1273 The type of tool being defined: `file_search`

1054 1274

1055 - `"auto"`1275 - `"file_search"`

1056 1276

1057 - `class ResponseFormatText: …`1277 - `file_search: Optional[FileSearch]`

1058 1278

1059 Default response format. Used to generate text responses.1279 Overrides for the file search tool.

1060 1280

1061 - `type: Literal["text"]`1281 - `max_num_results: Optional[int]`

1062 1282

1063 The type of response format being defined. Always `text`.1283 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.

1064 1284

1065 - `"text"`1285 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.

1066 1286

1067 - `class ResponseFormatJSONObject: …`1287 - `ranking_options: Optional[FileSearchRankingOptions]`

1068 1288

1069 JSON object response format. An older method of generating JSON responses.1289 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.

1070 Using `json_schema` is recommended for models that support it. Note that the

1071 model will not generate JSON without a system or user message instructing it

1072 to do so.

1073 1290

1074 - `type: Literal["json_object"]`1291 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

1075 1292

1076 The type of response format being defined. Always `json_object`.1293 - `score_threshold: float`

1077 1294

1078 - `"json_object"`1295 The score threshold for the file search. All values must be a floating point number between 0 and 1.

1079 1296

1080 - `class ResponseFormatJSONSchema: …`1297 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

1298

1299 The ranker to use for the file search. If not specified will use the `auto` ranker.

1300

1301 - `"auto"`

1302

1303 - `"default_2024_08_21"`

1304

1305 - `class FunctionTool: …`

1306

1307 - `function: FunctionDefinition`

1308

1309 - `name: str`

1310

1311 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.

1312

1313 - `description: Optional[str]`

1314

1315 A description of what the function does, used by the model to choose when and how to call the function.

1316

1317 - `parameters: Optional[FunctionParameters]`

1318

1319 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.

1320

1321 Omitting `parameters` defines a function with an empty parameter list.

1322

1323 - `strict: Optional[bool]`

1324

1325 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).

1326

1327 - `type: Literal["function"]`

1328

1329 The type of tool being defined: `function`

1330

1331 - `"function"`

1332

1333 - `response_format: Optional[AssistantResponseFormatOption]`

1334

1335 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`.

1336

1337 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).

1338

1339 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

1340

1341 **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.

1342

1343 - `Literal["auto"]`

1344

1345 `auto` is the default value

1346

1347 - `"auto"`

1348

1349 - `class ResponseFormatText: …`

1350

1351 Default response format. Used to generate text responses.

1352

1353 - `type: Literal["text"]`

1354

1355 The type of response format being defined. Always `text`.

1356

1357 - `"text"`

1358

1359 - `class ResponseFormatJSONObject: …`

1360

1361 JSON object response format. An older method of generating JSON responses.

1362 Using `json_schema` is recommended for models that support it. Note that the

1363 model will not generate JSON without a system or user message instructing it

1364 to do so.

1365

1366 - `type: Literal["json_object"]`

1367

1368 The type of response format being defined. Always `json_object`.

1369

1370 - `"json_object"`

1371

1372 - `class ResponseFormatJSONSchema: …`

1081 1373

1082 JSON Schema response format. Used to generate structured JSON responses.1374 JSON Schema response format. Used to generate structured JSON responses.

1083 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).1375 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).

1171 "name": "name",1463 "name": "name",

1172 "object": "assistant",1464 "object": "assistant",

1173 "tools": [1465 "tools": [

1174 {}1466 {

1467 "type": "code_interpreter"

1468 }

1175 ],1469 ],

1176 "response_format": "auto",1470 "response_format": "auto",

1177 "temperature": 1,1471 "temperature": 1,

1230 1524

1231**post** `/assistants/{assistant_id}`1525**post** `/assistants/{assistant_id}`

1232 1526

1233Modify assistant1527Modifies an assistant.

1234 1528

1235### Parameters1529### Parameters

1236 1530

1472 1766

1473 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.1767 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.

1474 1768

1475- `tools: Optional[Iterable[object]]`1769- `tools: Optional[Iterable[AssistantToolParam]]`

1476 1770

1477 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`.1771 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`.

1478 1772

1773 - `class CodeInterpreterTool: …`

1774

1775 - `type: Literal["code_interpreter"]`

1776

1777 The type of tool being defined: `code_interpreter`

1778

1779 - `"code_interpreter"`

1780

1781 - `class FileSearchTool: …`

1782

1783 - `type: Literal["file_search"]`

1784

1785 The type of tool being defined: `file_search`

1786

1787 - `"file_search"`

1788

1789 - `file_search: Optional[FileSearch]`

1790

1791 Overrides for the file search tool.

1792

1793 - `max_num_results: Optional[int]`

1794

1795 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.

1796

1797 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.

1798

1799 - `ranking_options: Optional[FileSearchRankingOptions]`

1800

1801 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.

1802

1803 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

1804

1805 - `score_threshold: float`

1806

1807 The score threshold for the file search. All values must be a floating point number between 0 and 1.

1808

1809 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

1810

1811 The ranker to use for the file search. If not specified will use the `auto` ranker.

1812

1813 - `"auto"`

1814

1815 - `"default_2024_08_21"`

1816

1817 - `class FunctionTool: …`

1818

1819 - `function: FunctionDefinition`

1820

1821 - `name: str`

1822

1823 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.

1824

1825 - `description: Optional[str]`

1826

1827 A description of what the function does, used by the model to choose when and how to call the function.

1828

1829 - `parameters: Optional[FunctionParameters]`

1830

1831 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.

1832

1833 Omitting `parameters` defines a function with an empty parameter list.

1834

1835 - `strict: Optional[bool]`

1836

1837 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).

1838

1839 - `type: Literal["function"]`

1840

1841 The type of tool being defined: `function`

1842

1843 - `"function"`

1844

1479- `top_p: Optional[float]`1845- `top_p: Optional[float]`

1480 1846

1481 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.1847 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.

1527 1893

1528 - `"assistant"`1894 - `"assistant"`

1529 1895

1530 - `tools: List[object]`1896 - `tools: List[AssistantTool]`

1531 1897

1532 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`.1898 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`.

1533 1899

1900 - `class CodeInterpreterTool: …`

1901

1902 - `type: Literal["code_interpreter"]`

1903

1904 The type of tool being defined: `code_interpreter`

1905

1906 - `"code_interpreter"`

1907

1908 - `class FileSearchTool: …`

1909

1910 - `type: Literal["file_search"]`

1911

1912 The type of tool being defined: `file_search`

1913

1914 - `"file_search"`

1915

1916 - `file_search: Optional[FileSearch]`

1917

1918 Overrides for the file search tool.

1919

1920 - `max_num_results: Optional[int]`

1921

1922 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.

1923

1924 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.

1925

1926 - `ranking_options: Optional[FileSearchRankingOptions]`

1927

1928 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.

1929

1930 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

1931

1932 - `score_threshold: float`

1933

1934 The score threshold for the file search. All values must be a floating point number between 0 and 1.

1935

1936 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

1937

1938 The ranker to use for the file search. If not specified will use the `auto` ranker.

1939

1940 - `"auto"`

1941

1942 - `"default_2024_08_21"`

1943

1944 - `class FunctionTool: …`

1945

1946 - `function: FunctionDefinition`

1947

1948 - `name: str`

1949

1950 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.

1951

1952 - `description: Optional[str]`

1953

1954 A description of what the function does, used by the model to choose when and how to call the function.

1955

1956 - `parameters: Optional[FunctionParameters]`

1957

1958 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.

1959

1960 Omitting `parameters` defines a function with an empty parameter list.

1961

1962 - `strict: Optional[bool]`

1963

1964 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).

1965

1966 - `type: Literal["function"]`

1967

1968 The type of tool being defined: `function`

1969

1970 - `"function"`

1971

1534 - `response_format: Optional[AssistantResponseFormatOption]`1972 - `response_format: Optional[AssistantResponseFormatOption]`

1535 1973

1536 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`.1974 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`.

1664 "name": "name",2102 "name": "name",

1665 "object": "assistant",2103 "object": "assistant",

1666 "tools": [2104 "tools": [

1667 {}2105 {

2106 "type": "code_interpreter"

2107 }

1668 ],2108 ],

1669 "response_format": "auto",2109 "response_format": "auto",

1670 "temperature": 1,2110 "temperature": 1,

1735 2175

1736**delete** `/assistants/{assistant_id}`2176**delete** `/assistants/{assistant_id}`

1737 2177

1738Delete assistant2178Delete an assistant.

1739 2179

1740### Parameters2180### Parameters

1741 2181

1845 2285

1846 - `"assistant"`2286 - `"assistant"`

1847 2287

1848 - `tools: List[object]`2288 - `tools: List[AssistantTool]`

1849 2289

1850 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`.2290 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`.

1851 2291

1852 - `response_format: Optional[AssistantResponseFormatOption]`2292 - `class CodeInterpreterTool: …`

1853 2293

1854 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`.2294 - `type: Literal["code_interpreter"]`

1855 2295

1856 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).2296 The type of tool being defined: `code_interpreter`

1857 2297

1858 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.2298 - `"code_interpreter"`

1859 2299

1860 **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.2300 - `class FileSearchTool: …`

1861 2301

1862 - `Literal["auto"]`2302 - `type: Literal["file_search"]`

1863 2303

1864 `auto` is the default value2304 The type of tool being defined: `file_search`

1865 2305

1866 - `"auto"`2306 - `"file_search"`

1867 2307

1868 - `class ResponseFormatText: …`2308 - `file_search: Optional[FileSearch]`

1869 2309

1870 Default response format. Used to generate text responses.2310 Overrides for the file search tool.

1871 2311

1872 - `type: Literal["text"]`2312 - `max_num_results: Optional[int]`

1873 2313

1874 The type of response format being defined. Always `text`.2314 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.

1875 2315

1876 - `"text"`2316 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.

1877 2317

1878 - `class ResponseFormatJSONObject: …`2318 - `ranking_options: Optional[FileSearchRankingOptions]`

1879 2319

1880 JSON object response format. An older method of generating JSON responses.2320 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.

1881 Using `json_schema` is recommended for models that support it. Note that the

1882 model will not generate JSON without a system or user message instructing it

1883 to do so.

1884 2321

1885 - `type: Literal["json_object"]`2322 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

1886 2323

1887 The type of response format being defined. Always `json_object`.2324 - `score_threshold: float`

2325

2326 The score threshold for the file search. All values must be a floating point number between 0 and 1.

2327

2328 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

2329

2330 The ranker to use for the file search. If not specified will use the `auto` ranker.

2331

2332 - `"auto"`

2333

2334 - `"default_2024_08_21"`

2335

2336 - `class FunctionTool: …`

2337

2338 - `function: FunctionDefinition`

2339

2340 - `name: str`

2341

2342 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.

2343

2344 - `description: Optional[str]`

2345

2346 A description of what the function does, used by the model to choose when and how to call the function.

2347

2348 - `parameters: Optional[FunctionParameters]`

2349

2350 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.

2351

2352 Omitting `parameters` defines a function with an empty parameter list.

2353

2354 - `strict: Optional[bool]`

2355

2356 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).

2357

2358 - `type: Literal["function"]`

2359

2360 The type of tool being defined: `function`

2361

2362 - `"function"`

2363

2364 - `response_format: Optional[AssistantResponseFormatOption]`

2365

2366 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`.

2367

2368 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).

2369

2370 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

2371

2372 **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.

2373

2374 - `Literal["auto"]`

2375

2376 `auto` is the default value

2377

2378 - `"auto"`

2379

2380 - `class ResponseFormatText: …`

2381

2382 Default response format. Used to generate text responses.

2383

2384 - `type: Literal["text"]`

2385

2386 The type of response format being defined. Always `text`.

2387

2388 - `"text"`

2389

2390 - `class ResponseFormatJSONObject: …`

2391

2392 JSON object response format. An older method of generating JSON responses.

2393 Using `json_schema` is recommended for models that support it. Note that the

2394 model will not generate JSON without a system or user message instructing it

2395 to do so.

2396

2397 - `type: Literal["json_object"]`

2398

2399 The type of response format being defined. Always `json_object`.

1888 2400

1889 - `"json_object"`2401 - `"json_object"`

1890 2402

2265 2777

2266 The Unix timestamp (in seconds) for when the run was started.2778 The Unix timestamp (in seconds) for when the run was started.

2267 2779

2268 - `status: object`2780 - `status: RunStatus`

2781

2782 The status of the run, which can be either `queued`, `in_progress`, `requires_action`, `cancelling`, `cancelled`, `failed`, `completed`, `incomplete`, or `expired`.

2783

2784 - `"queued"`

2785

2786 - `"in_progress"`

2787

2788 - `"requires_action"`

2789

2790 - `"cancelling"`

2791

2792 - `"cancelled"`

2793

2794 - `"failed"`

2795

2796 - `"completed"`

2797

2798 - `"incomplete"`

2799

2800 - `"expired"`

2269 2801

2270 - `thread_id: str`2802 - `thread_id: str`

2271 2803

2309 2841

2310 The name of the function to call.2842 The name of the function to call.

2311 2843

2312 - `tools: List[object]`2844 - `tools: List[AssistantTool]`

2313 2845

2314 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.2846 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

2315 2847

2848 - `class CodeInterpreterTool: …`

2849

2850 - `type: Literal["code_interpreter"]`

2851

2852 The type of tool being defined: `code_interpreter`

2853

2854 - `"code_interpreter"`

2855

2856 - `class FileSearchTool: …`

2857

2858 - `type: Literal["file_search"]`

2859

2860 The type of tool being defined: `file_search`

2861

2862 - `"file_search"`

2863

2864 - `file_search: Optional[FileSearch]`

2865

2866 Overrides for the file search tool.

2867

2868 - `max_num_results: Optional[int]`

2869

2870 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.

2871

2872 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.

2873

2874 - `ranking_options: Optional[FileSearchRankingOptions]`

2875

2876 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.

2877

2878 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

2879

2880 - `score_threshold: float`

2881

2882 The score threshold for the file search. All values must be a floating point number between 0 and 1.

2883

2884 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

2885

2886 The ranker to use for the file search. If not specified will use the `auto` ranker.

2887

2888 - `"auto"`

2889

2890 - `"default_2024_08_21"`

2891

2892 - `class FunctionTool: …`

2893

2894 - `function: FunctionDefinition`

2895

2896 - `name: str`

2897

2898 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.

2899

2900 - `description: Optional[str]`

2901

2902 A description of what the function does, used by the model to choose when and how to call the function.

2903

2904 - `parameters: Optional[FunctionParameters]`

2905

2906 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.

2907

2908 Omitting `parameters` defines a function with an empty parameter list.

2909

2910 - `strict: Optional[bool]`

2911

2912 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).

2913

2914 - `type: Literal["function"]`

2915

2916 The type of tool being defined: `function`

2917

2918 - `"function"`

2919

2316 - `truncation_strategy: Optional[TruncationStrategy]`2920 - `truncation_strategy: Optional[TruncationStrategy]`

2317 2921

2318 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.2922 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

2574 3178

2575 Details of the tool call.3179 Details of the tool call.

2576 3180

2577 - `tool_calls: List[object]`3181 - `tool_calls: List[ToolCall]`

2578 3182

2579 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`.3183 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`.

2580 3184

2581 - `type: Literal["tool_calls"]`3185 - `class CodeInterpreterToolCall: …`

~~2582~~

2583 Always `tool_calls`.

2584 3186

2585 - `"tool_calls"`3187 Details of the Code Interpreter tool call the run step was involved in.

2586 3188

2587 - `thread_id: str`3189 - `id: str`

2588 3190

2589 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.3191 The ID of the tool call.

2590 3192

2591 - `type: Literal["message_creation", "tool_calls"]`3193 - `code_interpreter: CodeInterpreter`

2592 3194

2593 The type of run step, which can be either `message_creation` or `tool_calls`.3195 The Code Interpreter tool call definition.

2594 3196

2595 - `"message_creation"`3197 - `input: str`

2596 3198

2597 - `"tool_calls"`3199 The input to the Code Interpreter tool call.

2598 3200

2599 - `usage: Optional[Usage]`3201 - `outputs: List[CodeInterpreterOutput]`

2600 3202

2601 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.3203 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.

2602 3204

2603 - `completion_tokens: int`3205 - `class CodeInterpreterOutputLogs: …`

2604 3206

2605 Number of completion tokens used over the course of the run step.3207 Text output from the Code Interpreter tool call as part of a run step.

2606 3208

2607 - `prompt_tokens: int`3209 - `logs: str`

2608 3210

2609 Number of prompt tokens used over the course of the run step.3211 The text output from the Code Interpreter tool call.

2610 3212

2611 - `total_tokens: int`3213 - `type: Literal["logs"]`

2612 3214

2613 Total number of tokens used (prompt + completion).3215 Always `logs`.

2614 3216

2615 - `event: Literal["thread.run.step.created"]`3217 - `"logs"`

2616 3218

2617 - `"thread.run.step.created"`3219 - `class CodeInterpreterOutputImage: …`

2618 3220

2619 - `class ThreadRunStepInProgress: …`3221 - `image: CodeInterpreterOutputImageImage`

2620 3222

2621 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) moves to an `in_progress` state.3223 - `file_id: str`

2622 3224

2623 - `data: RunStep`3225 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.

2624 3226

2625 Represents a step in execution of a run.3227 - `type: Literal["image"]`

2626 3228

2627 - `event: Literal["thread.run.step.in_progress"]`3229 Always `image`.

2628 3230

2629 - `"thread.run.step.in_progress"`3231 - `"image"`

2630 3232

2631 - `class ThreadRunStepDelta: …`3233 - `type: Literal["code_interpreter"]`

2632 3234

2633 Occurs when parts of a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) are being streamed.3235 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.

2634 3236

2635 - `data: RunStepDeltaEvent`3237 - `"code_interpreter"`

2636 3238

2637 Represents a run step delta i.e. any changed fields on a run step during streaming.3239 - `class FileSearchToolCall: …`

2638 3240

2639 - `id: str`3241 - `id: str`

2640 3242

2641 The identifier of the run step, which can be referenced in API endpoints.3243 The ID of the tool call object.

2642 3244

2643 - `delta: object`3245 - `file_search: FileSearch`

2644 3246

2645 - `object: Literal["thread.run.step.delta"]`3247 For now, this is always going to be an empty object.

2646 3248

2647 The object type, which is always `thread.run.step.delta`.3249 - `ranking_options: Optional[FileSearchRankingOptions]`

2648 3250

2649 - `"thread.run.step.delta"`3251 The ranking options for the file search.

2650 3252

2651 - `event: Literal["thread.run.step.delta"]`3253 - `ranker: Literal["auto", "default_2024_08_21"]`

2652 3254

2653 - `"thread.run.step.delta"`3255 The ranker to use for the file search. If not specified will use the `auto` ranker.

2654 3256

2655 - `class ThreadRunStepCompleted: …`3257 - `"auto"`

2656 3258

2657 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is completed.3259 - `"default_2024_08_21"`

2658 3260

2659 - `data: RunStep`3261 - `score_threshold: float`

2660 3262

2661 Represents a step in execution of a run.3263 The score threshold for the file search. All values must be a floating point number between 0 and 1.

2662 3264

2663 - `event: Literal["thread.run.step.completed"]`3265 - `results: Optional[List[FileSearchResult]]`

2664 3266

2665 - `"thread.run.step.completed"`3267 The results of the file search.

2666 3268

2667 - `class ThreadRunStepFailed: …`3269 - `file_id: str`

2668 3270

2669 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) fails.3271 The ID of the file that result was found in.

2670 3272

2671 - `data: RunStep`3273 - `file_name: str`

2672 3274

2673 Represents a step in execution of a run.3275 The name of the file that result was found in.

2674 3276

2675 - `event: Literal["thread.run.step.failed"]`3277 - `score: float`

2676 3278

2677 - `"thread.run.step.failed"`3279 The score of the result. All values must be a floating point number between 0 and 1.

2678 3280

2679 - `class ThreadRunStepCancelled: …`3281 - `content: Optional[List[FileSearchResultContent]]`

2680 3282

2681 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is cancelled.3283 The content of the result that was found. The content is only included if requested via the include query parameter.

2682 3284

2683 - `data: RunStep`3285 - `text: Optional[str]`

2684 3286

2685 Represents a step in execution of a run.3287 The text content of the file.

2686 3288

2687 - `event: Literal["thread.run.step.cancelled"]`3289 - `type: Optional[Literal["text"]]`

2688 3290

2689 - `"thread.run.step.cancelled"`3291 The type of the content.

2690 3292

2691 - `class ThreadRunStepExpired: …`3293 - `"text"`

2692 3294

2693 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) expires.3295 - `type: Literal["file_search"]`

2694 3296

2695 - `data: RunStep`3297 The type of tool call. This is always going to be `file_search` for this type of tool call.

2696 3298

2697 Represents a step in execution of a run.3299 - `"file_search"`

2698 3300

2699 - `event: Literal["thread.run.step.expired"]`3301 - `class FunctionToolCall: …`

2700 3302

2701 - `"thread.run.step.expired"`3303 - `id: str`

2702 3304

2703 - `class ThreadMessageCreated: …`3305 The ID of the tool call object.

2704 3306

2705 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.3307 - `function: Function`

2706 3308

2707 - `data: Message`3309 The definition of the function that was called.

2708 3310

2709 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).3311 - `arguments: str`

2710 3312

2711 - `id: str`3313 The arguments passed to the function.

2712 3314

2713 The identifier, which can be referenced in API endpoints.3315 - `name: str`

2714 3316

2715 - `assistant_id: Optional[str]`3317 The name of the function.

2716 3318

2717 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.3319 - `output: Optional[str]`

2718 3320

2719 - `attachments: Optional[List[Attachment]]`3321 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.

2720 3322

2721 A list of files attached to the message, and the tools they were added to.3323 - `type: Literal["function"]`

2722 3324

2723 - `file_id: Optional[str]`3325 The type of tool call. This is always going to be `function` for this type of tool call.

2724 3326

2725 The ID of the file to attach to the message.3327 - `"function"`

2726 3328

2727 - `tools: Optional[List[AttachmentTool]]`3329 - `type: Literal["tool_calls"]`

2728 3330

2729 The tools to add this file to.3331 Always `tool_calls`.

2730 3332

2731 - `class CodeInterpreterTool: …`3333 - `"tool_calls"`

2732 3334

2733 - `type: Literal["code_interpreter"]`3335 - `thread_id: str`

2734 3336

2735 The type of tool being defined: `code_interpreter`3337 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.

2736 3338

2737 - `"code_interpreter"`3339 - `type: Literal["message_creation", "tool_calls"]`

2738 3340

2739 - `class AttachmentToolAssistantToolsFileSearchTypeOnly: …`3341 The type of run step, which can be either `message_creation` or `tool_calls`.

2740 3342

2741 - `type: Literal["file_search"]`3343 - `"message_creation"`

2742 3344

2743 The type of tool being defined: `file_search`3345 - `"tool_calls"`

2744 3346

2745 - `"file_search"`3347 - `usage: Optional[Usage]`

2746 3348

2747 - `completed_at: Optional[int]`3349 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.

2748 3350

2749 The Unix timestamp (in seconds) for when the message was completed.3351 - `completion_tokens: int`

2750 3352

2751 - `content: List[object]`3353 Number of completion tokens used over the course of the run step.

2752 3354

2753 The content of the message in array of text and/or images.3355 - `prompt_tokens: int`

2754 3356

2755 - `created_at: int`3357 Number of prompt tokens used over the course of the run step.

2756 3358

2757 The Unix timestamp (in seconds) for when the message was created.3359 - `total_tokens: int`

2758 3360

2759 - `incomplete_at: Optional[int]`3361 Total number of tokens used (prompt + completion).

2760 3362

2761 The Unix timestamp (in seconds) for when the message was marked as incomplete.3363 - `event: Literal["thread.run.step.created"]`

2762 3364

2763 - `incomplete_details: Optional[IncompleteDetails]`3365 - `"thread.run.step.created"`

2764 3366

2765 On an incomplete message, details about why the message is incomplete.3367 - `class ThreadRunStepInProgress: …`

2766 3368

2767 - `reason: Literal["content_filter", "max_tokens", "run_cancelled", 2 more]`3369 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) moves to an `in_progress` state.

2768 3370

2769 The reason the message is incomplete.3371 - `data: RunStep`

2770 3372

2771 - `"content_filter"`3373 Represents a step in execution of a run.

2772 3374

2773 - `"max_tokens"`3375 - `event: Literal["thread.run.step.in_progress"]`

2774 3376

2775 - `"run_cancelled"`3377 - `"thread.run.step.in_progress"`

2776 3378

2777 - `"run_expired"`3379 - `class ThreadRunStepDelta: …`

2778 3380

2779 - `"run_failed"`3381 Occurs when parts of a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) are being streamed.

2780 3382

2781 - `metadata: Optional[Metadata]`3383 - `data: RunStepDeltaEvent`

2782 3384

2783 Set of 16 key-value pairs that can be attached to an object. This can be3385 Represents a run step delta i.e. any changed fields on a run step during streaming.

2784 useful for storing additional information about the object in a structured

2785 format, and querying for objects via API or the dashboard.

2786 3386

2787 Keys are strings with a maximum length of 64 characters. Values are strings3387 - `id: str`

2788 with a maximum length of 512 characters.

2789 3388

2790 - `object: Literal["thread.message"]`3389 The identifier of the run step, which can be referenced in API endpoints.

2791 3390

2792 The object type, which is always `thread.message`.3391 - `delta: RunStepDelta`

2793 3392

2794 - `"thread.message"`3393 The delta containing the fields that have changed on the run step.

2795 3394

2796 - `role: Literal["user", "assistant"]`3395 - `step_details: Optional[StepDetails]`

2797 3396

2798 The entity that produced the message. One of `user` or `assistant`.3397 The details of the run step.

2799 3398

2800 - `"user"`3399 - `class RunStepDeltaMessageDelta: …`

2801 3400

2802 - `"assistant"`3401 Details of the message creation by the run step.

2803 3402

2804 - `run_id: Optional[str]`3403 - `type: Literal["message_creation"]`

2805 3404

2806 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.3405 Always `message_creation`.

2807 3406

2808 - `status: Literal["in_progress", "incomplete", "completed"]`3407 - `"message_creation"`

2809 3408

2810 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.3409 - `message_creation: Optional[MessageCreation]`

2811 3410

2812 - `"in_progress"`3411 - `message_id: Optional[str]`

2813 3412

2814 - `"incomplete"`3413 The ID of the message that was created by this run step.

2815 3414

2816 - `"completed"`3415 - `class ToolCallDeltaObject: …`

2817 3416

2818 - `thread_id: str`3417 Details of the tool call.

2819 3418

2820 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.3419 - `type: Literal["tool_calls"]`

2821 3420

2822 - `event: Literal["thread.message.created"]`3421 Always `tool_calls`.

2823 3422

2824 - `"thread.message.created"`3423 - `"tool_calls"`

2825 3424

2826 - `class ThreadMessageInProgress: …`3425 - `tool_calls: Optional[List[ToolCallDelta]]`

2827 3426

2828 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) moves to an `in_progress` state.3427 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`.

2829 3428

2830 - `data: Message`3429 - `class CodeInterpreterToolCallDelta: …`

2831 3430

2832 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).3431 Details of the Code Interpreter tool call the run step was involved in.

2833 3432

2834 - `event: Literal["thread.message.in_progress"]`3433 - `index: int`

2835 3434

2836 - `"thread.message.in_progress"`3435 The index of the tool call in the tool calls array.

2837 3436

2838 - `class ThreadMessageDelta: …`3437 - `type: Literal["code_interpreter"]`

2839 3438

2840 Occurs when parts of a [Message](https://platform.openai.com/docs/api-reference/messages/object) are being streamed.3439 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.

2841 3440

2842 - `data: MessageDeltaEvent`3441 - `"code_interpreter"`

2843 3442

2844 Represents a message delta i.e. any changed fields on a message during streaming.3443 - `id: Optional[str]`

2845 3444

2846 - `id: str`3445 The ID of the tool call.

2847 3446

2848 The identifier of the message, which can be referenced in API endpoints.3447 - `code_interpreter: Optional[CodeInterpreter]`

2849 3448

2850 - `delta: MessageDelta`3449 The Code Interpreter tool call definition.

2851 3450

2852 The delta containing the fields that have changed on the Message.3451 - `input: Optional[str]`

2853 3452

2854 - `content: Optional[List[object]]`3453 The input to the Code Interpreter tool call.

2855 3454

2856 The content of the message in array of text and/or images.3455 - `outputs: Optional[List[CodeInterpreterOutput]]`

2857 3456

2858 - `role: Optional[Literal["user", "assistant"]]`3457 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.

2859 3458

2860 The entity that produced the message. One of `user` or `assistant`.3459 - `class CodeInterpreterLogs: …`

2861 3460

2862 - `"user"`3461 Text output from the Code Interpreter tool call as part of a run step.

2863 3462

2864 - `"assistant"`3463 - `index: int`

2865 3464

2866 - `object: Literal["thread.message.delta"]`3465 The index of the output in the outputs array.

2867 3466

2868 The object type, which is always `thread.message.delta`.3467 - `type: Literal["logs"]`

2869 3468

2870 - `"thread.message.delta"`3469 Always `logs`.

2871 3470

2872 - `event: Literal["thread.message.delta"]`3471 - `"logs"`

2873 3472

2874 - `"thread.message.delta"`3473 - `logs: Optional[str]`

2875 3474

2876 - `class ThreadMessageCompleted: …`3475 The text output from the Code Interpreter tool call.

2877 3476

2878 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is completed.3477 - `class CodeInterpreterOutputImage: …`

2879 3478

2880 - `data: Message`3479 - `index: int`

2881 3480

2882 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).3481 The index of the output in the outputs array.

2883 3482

2884 - `event: Literal["thread.message.completed"]`3483 - `type: Literal["image"]`

2885 3484

2886 - `"thread.message.completed"`3485 Always `image`.

2887 3486

2888 - `class ThreadMessageIncomplete: …`3487 - `"image"`

2889 3488

2890 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) ends before it is completed.3489 - `image: Optional[Image]`

2891 3490

2892 - `data: Message`3491 - `file_id: Optional[str]`

2893 3492

2894 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).3493 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.

2895 3494

2896 - `event: Literal["thread.message.incomplete"]`3495 - `class FileSearchToolCallDelta: …`

2897 3496

2898 - `"thread.message.incomplete"`3497 - `file_search: object`

2899 3498

2900 - `class ErrorEvent: …`3499 For now, this is always going to be an empty object.

2901 3500

2902 Occurs when an [error](https://platform.openai.com/docs/guides/error-codes#api-errors) occurs. This can happen due to an internal server error or a timeout.3501 - `index: int`

2903 3502

2904 - `data: ErrorObject`3503 The index of the tool call in the tool calls array.

2905 3504

2906 - `code: Optional[str]`3505 - `type: Literal["file_search"]`

2907 3506

2908 - `message: str`3507 The type of tool call. This is always going to be `file_search` for this type of tool call.

2909 3508

2910 - `param: Optional[str]`3509 - `"file_search"`

2911 3510

2912 - `type: str`3511 - `id: Optional[str]`

2913 3512

2914 - `event: Literal["error"]`3513 The ID of the tool call object.

2915 3514

2916 - `"error"`3515 - `class FunctionToolCallDelta: …`

2917 3516

2918### Assistant Tool3517 - `index: int`

2919 3518

2920- `object`3519 The index of the tool call in the tool calls array.

2921 3520

2922### Code Interpreter Tool3521 - `type: Literal["function"]`

2923 3522

2924- `class CodeInterpreterTool: …`3523 The type of tool call. This is always going to be `function` for this type of tool call.

2925 3524

2926 - `type: Literal["code_interpreter"]`3525 - `"function"`

2927 3526

2928 The type of tool being defined: `code_interpreter`3527 - `id: Optional[str]`

2929 3528

2930 - `"code_interpreter"`3529 The ID of the tool call object.

2931 3530

2932### File Search Tool3531 - `function: Optional[Function]`

2933 3532

2934- `class FileSearchTool: …`3533 The definition of the function that was called.

2935 3534

2936 - `type: Literal["file_search"]`3535 - `arguments: Optional[str]`

2937 3536

2938 The type of tool being defined: `file_search`3537 The arguments passed to the function.

2939 3538

2940 - `"file_search"`3539 - `name: Optional[str]`

2941 3540

2942 - `file_search: Optional[FileSearch]`3541 The name of the function.

2943 3542

2944 Overrides for the file search tool.3543 - `output: Optional[str]`

2945 3544

2946 - `max_num_results: Optional[int]`3545 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.

2947 3546

2948 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.3547 - `object: Literal["thread.run.step.delta"]`

2949 3548

2950 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.3549 The object type, which is always `thread.run.step.delta`.

2951 3550

2952 - `ranking_options: Optional[FileSearchRankingOptions]`3551 - `"thread.run.step.delta"`

2953 3552

2954 The ranking options for the file search. If not specified, the file search tool will use the `auto` ranker and a score_threshold of 0.3553 - `event: Literal["thread.run.step.delta"]`

2955 3554

2956 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.3555 - `"thread.run.step.delta"`

2957 3556

2958 - `score_threshold: float`3557 - `class ThreadRunStepCompleted: …`

2959 3558

2960 The score threshold for the file search. All values must be a floating point number between 0 and 1.3559 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is completed.

2961 3560

2962 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`3561 - `data: RunStep`

2963 3562

2964 The ranker to use for the file search. If not specified will use the `auto` ranker.3563 Represents a step in execution of a run.

2965 3564

2966 - `"auto"`3565 - `event: Literal["thread.run.step.completed"]`

2967 3566

2968 - `"default_2024_08_21"`3567 - `"thread.run.step.completed"`

2969 3568

2970### Function Tool3569 - `class ThreadRunStepFailed: …`

2971 3570

2972- `class FunctionTool: …`3571 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) fails.

2973 3572

2974 - `function: FunctionDefinition`3573 - `data: RunStep`

2975 3574

2976 - `name: str`3575 Represents a step in execution of a run.

2977 3576

2978 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 64.3577 - `event: Literal["thread.run.step.failed"]`

2979 3578

2980 - `description: Optional[str]`3579 - `"thread.run.step.failed"`

2981 3580

2982 A description of what the function does, used by the model to choose when and how to call the function.3581 - `class ThreadRunStepCancelled: …`

2983 3582

2984 - `parameters: Optional[FunctionParameters]`3583 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is cancelled.

2985 3584

2986 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.3585 - `data: RunStep`

2987 3586

2988 Omitting `parameters` defines a function with an empty parameter list.3587 Represents a step in execution of a run.

2989 3588

2990 - `strict: Optional[bool]`3589 - `event: Literal["thread.run.step.cancelled"]`

2991 3590

2992 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).3591 - `"thread.run.step.cancelled"`

2993 3592

2994 - `type: Literal["function"]`3593 - `class ThreadRunStepExpired: …`

2995 3594

2996 The type of tool being defined: `function`3595 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) expires.

2997 3596

2998 - `"function"`3597 - `data: RunStep`

2999 3598

3000### Message Stream Event3599 Represents a step in execution of a run.

3001 3600

3002- `MessageStreamEvent`3601 - `event: Literal["thread.run.step.expired"]`

3003 3602

3004 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.3603 - `"thread.run.step.expired"`

3005 3604

3006 - `class ThreadMessageCreated: …`3605 - `class ThreadMessageCreated: …`

3007 3606

3033 3632

3034 - `class CodeInterpreterTool: …`3633 - `class CodeInterpreterTool: …`

3035 3634

3036 - `type: Literal["code_interpreter"]`

~~3037~~

3038 The type of tool being defined: `code_interpreter`

~~3039~~

3040 - `"code_interpreter"`

~~3041~~

3042 - `class AttachmentToolAssistantToolsFileSearchTypeOnly: …`3635 - `class AttachmentToolAssistantToolsFileSearchTypeOnly: …`

3043 3636

3044 - `type: Literal["file_search"]`3637 - `type: Literal["file_search"]`

3051 3644

3052 The Unix timestamp (in seconds) for when the message was completed.3645 The Unix timestamp (in seconds) for when the message was completed.

3053 3646

3054 - `content: List[object]`3647 - `content: List[MessageContent]`

3055 3648

3056 The content of the message in array of text and/or images.3649 The content of the message in array of text and/or images.

3057 3650

3058 - `created_at: int`3651 - `class ImageFileContentBlock: …`

3059 3652

3060 The Unix timestamp (in seconds) for when the message was created.3653 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

3061 3654

3062 - `incomplete_at: Optional[int]`3655 - `image_file: ImageFile`

3063 3656

3064 The Unix timestamp (in seconds) for when the message was marked as incomplete.3657 - `file_id: str`

3065 3658

3066 - `incomplete_details: Optional[IncompleteDetails]`3659 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.

3067 3660

3068 On an incomplete message, details about why the message is incomplete.3661 - `detail: Optional[Literal["auto", "low", "high"]]`

3069 3662

3070 - `reason: Literal["content_filter", "max_tokens", "run_cancelled", 2 more]`3663 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`.

3071 3664

3072 The reason the message is incomplete.3665 - `"auto"`

3073 3666

3074 - `"content_filter"`3667 - `"low"`

3075 3668

3076 - `"max_tokens"`3669 - `"high"`

3077 3670

3078 - `"run_cancelled"`3671 - `type: Literal["image_file"]`

3079 3672

3080 - `"run_expired"`3673 Always `image_file`.

3081 3674

3082 - `"run_failed"`3675 - `"image_file"`

3083 3676

3084 - `metadata: Optional[Metadata]`3677 - `class ImageURLContentBlock: …`

3085 3678

3086 Set of 16 key-value pairs that can be attached to an object. This can be3679 References an image URL in the content of a message.

3087 useful for storing additional information about the object in a structured

3088 format, and querying for objects via API or the dashboard.

3089 3680

3090 Keys are strings with a maximum length of 64 characters. Values are strings3681 - `image_url: ImageURL`

3091 with a maximum length of 512 characters.

3092 3682

3093 - `object: Literal["thread.message"]`3683 - `url: str`

3094 3684

3095 The object type, which is always `thread.message`.3685 The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

3096 3686

3097 - `"thread.message"`3687 - `detail: Optional[Literal["auto", "low", "high"]]`

3098 3688

3099 - `role: Literal["user", "assistant"]`3689 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`. Default value is `auto`

3100 3690

3101 The entity that produced the message. One of `user` or `assistant`.3691 - `"auto"`

3102 3692

3103 - `"user"`3693 - `"low"`

3104 3694

3105 - `"assistant"`3695 - `"high"`

3106 3696

3107 - `run_id: Optional[str]`3697 - `type: Literal["image_url"]`

3108 3698

3109 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.3699 The type of the content part.

3110 3700

3111 - `status: Literal["in_progress", "incomplete", "completed"]`3701 - `"image_url"`

3112 3702

3113 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.3703 - `class TextContentBlock: …`

3114 3704

3115 - `"in_progress"`3705 The text content that is part of a message.

3706

3707 - `text: Text`

3708

3709 - `annotations: List[Annotation]`

3710

3711 - `class FileCitationAnnotation: …`

3712

3713 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.

3714

3715 - `end_index: int`

3716

3717 - `file_citation: FileCitation`

3718

3719 - `file_id: str`

3720

3721 The ID of the specific File the citation is from.

3722

3723 - `start_index: int`

3724

3725 - `text: str`

3726

3727 The text in the message content that needs to be replaced.

3728

3729 - `type: Literal["file_citation"]`

3730

3731 Always `file_citation`.

3732

3733 - `"file_citation"`

3734

3735 - `class FilePathAnnotation: …`

3736

3737 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.

3738

3739 - `end_index: int`

3740

3741 - `file_path: FilePath`

3742

3743 - `file_id: str`

3744

3745 The ID of the file that was generated.

3746

3747 - `start_index: int`

3748

3749 - `text: str`

3750

3751 The text in the message content that needs to be replaced.

3752

3753 - `type: Literal["file_path"]`

3754

3755 Always `file_path`.

3756

3757 - `"file_path"`

3758

3759 - `value: str`

3760

3761 The data that makes up the text.

3762

3763 - `type: Literal["text"]`

3764

3765 Always `text`.

3766

3767 - `"text"`

3768

3769 - `class RefusalContentBlock: …`

3770

3771 The refusal content generated by the assistant.

3772

3773 - `refusal: str`

3774

3775 - `type: Literal["refusal"]`

3776

3777 Always `refusal`.

3778

3779 - `"refusal"`

3780

3781 - `created_at: int`

3782

3783 The Unix timestamp (in seconds) for when the message was created.

3784

3785 - `incomplete_at: Optional[int]`

3786

3787 The Unix timestamp (in seconds) for when the message was marked as incomplete.

3788

3789 - `incomplete_details: Optional[IncompleteDetails]`

3790

3791 On an incomplete message, details about why the message is incomplete.

3792

3793 - `reason: Literal["content_filter", "max_tokens", "run_cancelled", 2 more]`

3794

3795 The reason the message is incomplete.

3796

3797 - `"content_filter"`

3798

3799 - `"max_tokens"`

3800

3801 - `"run_cancelled"`

3802

3803 - `"run_expired"`

3804

3805 - `"run_failed"`

3806

3807 - `metadata: Optional[Metadata]`

3808

3809 Set of 16 key-value pairs that can be attached to an object. This can be

3810 useful for storing additional information about the object in a structured

3811 format, and querying for objects via API or the dashboard.

3812

3813 Keys are strings with a maximum length of 64 characters. Values are strings

3814 with a maximum length of 512 characters.

3815

3816 - `object: Literal["thread.message"]`

3817

3818 The object type, which is always `thread.message`.

3819

3820 - `"thread.message"`

3821

3822 - `role: Literal["user", "assistant"]`

3823

3824 The entity that produced the message. One of `user` or `assistant`.

3825

3826 - `"user"`

3827

3828 - `"assistant"`

3829

3830 - `run_id: Optional[str]`

3831

3832 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.

3833

3834 - `status: Literal["in_progress", "incomplete", "completed"]`

3835

3836 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

3837

3838 - `"in_progress"`

3116 3839

3117 - `"incomplete"`3840 - `"incomplete"`

3118 3841

3150 3873

3151 The identifier of the message, which can be referenced in API endpoints.3874 The identifier of the message, which can be referenced in API endpoints.

3152 3875

3153 - `delta: MessageDelta`3876 - `delta: MessageDelta`

3877

3878 The delta containing the fields that have changed on the Message.

3879

3880 - `content: Optional[List[MessageContentDelta]]`

3881

3882 The content of the message in array of text and/or images.

3883

3884 - `class ImageFileDeltaBlock: …`

3885

3886 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

3887

3888 - `index: int`

3889

3890 The index of the content part in the message.

3891

3892 - `type: Literal["image_file"]`

3893

3894 Always `image_file`.

3895

3896 - `"image_file"`

3897

3898 - `image_file: Optional[ImageFileDelta]`

3899

3900 - `detail: Optional[Literal["auto", "low", "high"]]`

3901

3902 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`.

3903

3904 - `"auto"`

3905

3906 - `"low"`

3907

3908 - `"high"`

3909

3910 - `file_id: Optional[str]`

3911

3912 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.

3913

3914 - `class TextDeltaBlock: …`

3915

3916 The text content that is part of a message.

3917

3918 - `index: int`

3919

3920 The index of the content part in the message.

3921

3922 - `type: Literal["text"]`

3923

3924 Always `text`.

3925

3926 - `"text"`

3927

3928 - `text: Optional[TextDelta]`

3929

3930 - `annotations: Optional[List[AnnotationDelta]]`

3931

3932 - `class FileCitationDeltaAnnotation: …`

3933

3934 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.

3935

3936 - `index: int`

3937

3938 The index of the annotation in the text content part.

3939

3940 - `type: Literal["file_citation"]`

3941

3942 Always `file_citation`.

3943

3944 - `"file_citation"`

3945

3946 - `end_index: Optional[int]`

3947

3948 - `file_citation: Optional[FileCitation]`

3949

3950 - `file_id: Optional[str]`

3951

3952 The ID of the specific File the citation is from.

3953

3954 - `quote: Optional[str]`

3955

3956 The specific quote in the file.

3957

3958 - `start_index: Optional[int]`

3959

3960 - `text: Optional[str]`

3961

3962 The text in the message content that needs to be replaced.

3963

3964 - `class FilePathDeltaAnnotation: …`

3965

3966 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.

3967

3968 - `index: int`

3969

3970 The index of the annotation in the text content part.

3971

3972 - `type: Literal["file_path"]`

3973

3974 Always `file_path`.

3975

3976 - `"file_path"`

3977

3978 - `end_index: Optional[int]`

3979

3980 - `file_path: Optional[FilePath]`

3981

3982 - `file_id: Optional[str]`

3983

3984 The ID of the file that was generated.

3985

3986 - `start_index: Optional[int]`

3987

3988 - `text: Optional[str]`

3989

3990 The text in the message content that needs to be replaced.

3991

3992 - `value: Optional[str]`

3993

3994 The data that makes up the text.

3995

3996 - `class RefusalDeltaBlock: …`

3997

3998 The refusal content that is part of a message.

3999

4000 - `index: int`

4001

4002 The index of the refusal part in the message.

4003

4004 - `type: Literal["refusal"]`

4005

4006 Always `refusal`.

4007

4008 - `"refusal"`

4009

4010 - `refusal: Optional[str]`

4011

4012 - `class ImageURLDeltaBlock: …`

4013

4014 References an image URL in the content of a message.

4015

4016 - `index: int`

4017

4018 The index of the content part in the message.

4019

4020 - `type: Literal["image_url"]`

4021

4022 Always `image_url`.

4023

4024 - `"image_url"`

4025

4026 - `image_url: Optional[ImageURLDelta]`

4027

4028 - `detail: Optional[Literal["auto", "low", "high"]]`

4029

4030 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`.

4031

4032 - `"auto"`

4033

4034 - `"low"`

4035

4036 - `"high"`

4037

4038 - `url: Optional[str]`

4039

4040 The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

4041

4042 - `role: Optional[Literal["user", "assistant"]]`

4043

4044 The entity that produced the message. One of `user` or `assistant`.

4045

4046 - `"user"`

4047

4048 - `"assistant"`

4049

4050 - `object: Literal["thread.message.delta"]`

4051

4052 The object type, which is always `thread.message.delta`.

4053

4054 - `"thread.message.delta"`

4055

4056 - `event: Literal["thread.message.delta"]`

4057

4058 - `"thread.message.delta"`

4059

4060 - `class ThreadMessageCompleted: …`

4061

4062 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is completed.

4063

4064 - `data: Message`

4065

4066 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4067

4068 - `event: Literal["thread.message.completed"]`

4069

4070 - `"thread.message.completed"`

4071

4072 - `class ThreadMessageIncomplete: …`

4073

4074 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) ends before it is completed.

4075

4076 - `data: Message`

4077

4078 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4079

4080 - `event: Literal["thread.message.incomplete"]`

4081

4082 - `"thread.message.incomplete"`

4083

4084 - `class ErrorEvent: …`

4085

4086 Occurs when an [error](https://platform.openai.com/docs/guides/error-codes#api-errors) occurs. This can happen due to an internal server error or a timeout.

4087

4088 - `data: ErrorObject`

4089

4090 - `code: Optional[str]`

4091

4092 - `message: str`

4093

4094 - `param: Optional[str]`

4095

4096 - `type: str`

4097

4098 - `event: Literal["error"]`

4099

4100 - `"error"`

4101

4102### Assistant Tool

4103

4104- `AssistantTool`

4105

4106 - `class CodeInterpreterTool: …`

4107

4108 - `type: Literal["code_interpreter"]`

4109

4110 The type of tool being defined: `code_interpreter`

4111

4112 - `"code_interpreter"`

4113

4114 - `class FileSearchTool: …`

4115

4116 - `type: Literal["file_search"]`

4117

4118 The type of tool being defined: `file_search`

4119

4120 - `"file_search"`

4121

4122 - `file_search: Optional[FileSearch]`

4123

4124 Overrides for the file search tool.

4125

4126 - `max_num_results: Optional[int]`

4127

4128 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.

4129

4130 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.

4131

4132 - `ranking_options: Optional[FileSearchRankingOptions]`

4133

4134 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.

4135

4136 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

4137

4138 - `score_threshold: float`

4139

4140 The score threshold for the file search. All values must be a floating point number between 0 and 1.

4141

4142 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

4143

4144 The ranker to use for the file search. If not specified will use the `auto` ranker.

4145

4146 - `"auto"`

4147

4148 - `"default_2024_08_21"`

4149

4150 - `class FunctionTool: …`

4151

4152 - `function: FunctionDefinition`

4153

4154 - `name: str`

4155

4156 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.

4157

4158 - `description: Optional[str]`

4159

4160 A description of what the function does, used by the model to choose when and how to call the function.

4161

4162 - `parameters: Optional[FunctionParameters]`

4163

4164 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.

4165

4166 Omitting `parameters` defines a function with an empty parameter list.

4167

4168 - `strict: Optional[bool]`

4169

4170 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).

4171

4172 - `type: Literal["function"]`

4173

4174 The type of tool being defined: `function`

4175

4176 - `"function"`

4177

4178### Code Interpreter Tool

4179

4180- `class CodeInterpreterTool: …`

4181

4182 - `type: Literal["code_interpreter"]`

4183

4184 The type of tool being defined: `code_interpreter`

4185

4186 - `"code_interpreter"`

4187

4188### File Search Tool

4189

4190- `class FileSearchTool: …`

4191

4192 - `type: Literal["file_search"]`

4193

4194 The type of tool being defined: `file_search`

4195

4196 - `"file_search"`

4197

4198 - `file_search: Optional[FileSearch]`

4199

4200 Overrides for the file search tool.

4201

4202 - `max_num_results: Optional[int]`

4203

4204 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.

4205

4206 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.

4207

4208 - `ranking_options: Optional[FileSearchRankingOptions]`

4209

4210 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.

4211

4212 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

4213

4214 - `score_threshold: float`

4215

4216 The score threshold for the file search. All values must be a floating point number between 0 and 1.

4217

4218 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

4219

4220 The ranker to use for the file search. If not specified will use the `auto` ranker.

4221

4222 - `"auto"`

4223

4224 - `"default_2024_08_21"`

4225

4226### Function Tool

4227

4228- `class FunctionTool: …`

4229

4230 - `function: FunctionDefinition`

4231

4232 - `name: str`

4233

4234 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.

4235

4236 - `description: Optional[str]`

4237

4238 A description of what the function does, used by the model to choose when and how to call the function.

4239

4240 - `parameters: Optional[FunctionParameters]`

4241

4242 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.

4243

4244 Omitting `parameters` defines a function with an empty parameter list.

4245

4246 - `strict: Optional[bool]`

4247

4248 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).

4249

4250 - `type: Literal["function"]`

4251

4252 The type of tool being defined: `function`

4253

4254 - `"function"`

4255

4256### Message Stream Event

4257

4258- `MessageStreamEvent`

4259

4260 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.

4261

4262 - `class ThreadMessageCreated: …`

4263

4264 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.

4265

4266 - `data: Message`

4267

4268 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4269

4270 - `id: str`

4271

4272 The identifier, which can be referenced in API endpoints.

4273

4274 - `assistant_id: Optional[str]`

4275

4276 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.

4277

4278 - `attachments: Optional[List[Attachment]]`

4279

4280 A list of files attached to the message, and the tools they were added to.

4281

4282 - `file_id: Optional[str]`

4283

4284 The ID of the file to attach to the message.

4285

4286 - `tools: Optional[List[AttachmentTool]]`

4287

4288 The tools to add this file to.

4289

4290 - `class CodeInterpreterTool: …`

4291

4292 - `type: Literal["code_interpreter"]`

4293

4294 The type of tool being defined: `code_interpreter`

4295

4296 - `"code_interpreter"`

4297

4298 - `class AttachmentToolAssistantToolsFileSearchTypeOnly: …`

4299

4300 - `type: Literal["file_search"]`

4301

4302 The type of tool being defined: `file_search`

4303

4304 - `"file_search"`

4305

4306 - `completed_at: Optional[int]`

4307

4308 The Unix timestamp (in seconds) for when the message was completed.

4309

4310 - `content: List[MessageContent]`

4311

4312 The content of the message in array of text and/or images.

4313

4314 - `class ImageFileContentBlock: …`

4315

4316 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

4317

4318 - `image_file: ImageFile`

4319

4320 - `file_id: str`

4321

4322 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.

4323

4324 - `detail: Optional[Literal["auto", "low", "high"]]`

4325

4326 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`.

4327

4328 - `"auto"`

4329

4330 - `"low"`

4331

4332 - `"high"`

4333

4334 - `type: Literal["image_file"]`

4335

4336 Always `image_file`.

4337

4338 - `"image_file"`

4339

4340 - `class ImageURLContentBlock: …`

4341

4342 References an image URL in the content of a message.

4343

4344 - `image_url: ImageURL`

4345

4346 - `url: str`

4347

4348 The external URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

4349

4350 - `detail: Optional[Literal["auto", "low", "high"]]`

4351

4352 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`. Default value is `auto`

4353

4354 - `"auto"`

4355

4356 - `"low"`

4357

4358 - `"high"`

4359

4360 - `type: Literal["image_url"]`

4361

4362 The type of the content part.

4363

4364 - `"image_url"`

4365

4366 - `class TextContentBlock: …`

4367

4368 The text content that is part of a message.

4369

4370 - `text: Text`

4371

4372 - `annotations: List[Annotation]`

4373

4374 - `class FileCitationAnnotation: …`

4375

4376 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.

4377

4378 - `end_index: int`

4379

4380 - `file_citation: FileCitation`

4381

4382 - `file_id: str`

4383

4384 The ID of the specific File the citation is from.

4385

4386 - `start_index: int`

4387

4388 - `text: str`

4389

4390 The text in the message content that needs to be replaced.

4391

4392 - `type: Literal["file_citation"]`

4393

4394 Always `file_citation`.

4395

4396 - `"file_citation"`

4397

4398 - `class FilePathAnnotation: …`

4399

4400 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.

4401

4402 - `end_index: int`

4403

4404 - `file_path: FilePath`

4405

4406 - `file_id: str`

4407

4408 The ID of the file that was generated.

4409

4410 - `start_index: int`

4411

4412 - `text: str`

4413

4414 The text in the message content that needs to be replaced.

4415

4416 - `type: Literal["file_path"]`

4417

4418 Always `file_path`.

4419

4420 - `"file_path"`

4421

4422 - `value: str`

4423

4424 The data that makes up the text.

4425

4426 - `type: Literal["text"]`

4427

4428 Always `text`.

4429

4430 - `"text"`

4431

4432 - `class RefusalContentBlock: …`

4433

4434 The refusal content generated by the assistant.

4435

4436 - `refusal: str`

4437

4438 - `type: Literal["refusal"]`

4439

4440 Always `refusal`.

4441

4442 - `"refusal"`

4443

4444 - `created_at: int`

4445

4446 The Unix timestamp (in seconds) for when the message was created.

4447

4448 - `incomplete_at: Optional[int]`

4449

4450 The Unix timestamp (in seconds) for when the message was marked as incomplete.

4451

4452 - `incomplete_details: Optional[IncompleteDetails]`

4453

4454 On an incomplete message, details about why the message is incomplete.

4455

4456 - `reason: Literal["content_filter", "max_tokens", "run_cancelled", 2 more]`

4457

4458 The reason the message is incomplete.

4459

4460 - `"content_filter"`

4461

4462 - `"max_tokens"`

4463

4464 - `"run_cancelled"`

4465

4466 - `"run_expired"`

4467

4468 - `"run_failed"`

4469

4470 - `metadata: Optional[Metadata]`

4471

4472 Set of 16 key-value pairs that can be attached to an object. This can be

4473 useful for storing additional information about the object in a structured

4474 format, and querying for objects via API or the dashboard.

4475

4476 Keys are strings with a maximum length of 64 characters. Values are strings

4477 with a maximum length of 512 characters.

4478

4479 - `object: Literal["thread.message"]`

4480

4481 The object type, which is always `thread.message`.

4482

4483 - `"thread.message"`

4484

4485 - `role: Literal["user", "assistant"]`

4486

4487 The entity that produced the message. One of `user` or `assistant`.

4488

4489 - `"user"`

4490

4491 - `"assistant"`

4492

4493 - `run_id: Optional[str]`

4494

4495 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.

4496

4497 - `status: Literal["in_progress", "incomplete", "completed"]`

4498

4499 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

4500

4501 - `"in_progress"`

4502

4503 - `"incomplete"`

4504

4505 - `"completed"`

4506

4507 - `thread_id: str`

4508

4509 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.

4510

4511 - `event: Literal["thread.message.created"]`

4512

4513 - `"thread.message.created"`

4514

4515 - `class ThreadMessageInProgress: …`

4516

4517 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) moves to an `in_progress` state.

4518

4519 - `data: Message`

4520

4521 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4522

4523 - `event: Literal["thread.message.in_progress"]`

4524

4525 - `"thread.message.in_progress"`

4526

4527 - `class ThreadMessageDelta: …`

4528

4529 Occurs when parts of a [Message](https://platform.openai.com/docs/api-reference/messages/object) are being streamed.

4530

4531 - `data: MessageDeltaEvent`

4532

4533 Represents a message delta i.e. any changed fields on a message during streaming.

4534

4535 - `id: str`

4536

4537 The identifier of the message, which can be referenced in API endpoints.

4538

4539 - `delta: MessageDelta`

4540

4541 The delta containing the fields that have changed on the Message.

4542

4543 - `content: Optional[List[MessageContentDelta]]`

4544

4545 The content of the message in array of text and/or images.

4546

4547 - `class ImageFileDeltaBlock: …`

4548

4549 References an image [File](https://platform.openai.com/docs/api-reference/files) in the content of a message.

4550

4551 - `index: int`

4552

4553 The index of the content part in the message.

4554

4555 - `type: Literal["image_file"]`

4556

4557 Always `image_file`.

4558

4559 - `"image_file"`

4560

4561 - `image_file: Optional[ImageFileDelta]`

4562

4563 - `detail: Optional[Literal["auto", "low", "high"]]`

4564

4565 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`.

4566

4567 - `"auto"`

4568

4569 - `"low"`

4570

4571 - `"high"`

4572

4573 - `file_id: Optional[str]`

4574

4575 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.

4576

4577 - `class TextDeltaBlock: …`

4578

4579 The text content that is part of a message.

4580

4581 - `index: int`

4582

4583 The index of the content part in the message.

4584

4585 - `type: Literal["text"]`

4586

4587 Always `text`.

4588

4589 - `"text"`

4590

4591 - `text: Optional[TextDelta]`

4592

4593 - `annotations: Optional[List[AnnotationDelta]]`

4594

4595 - `class FileCitationDeltaAnnotation: …`

4596

4597 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.

4598

4599 - `index: int`

4600

4601 The index of the annotation in the text content part.

4602

4603 - `type: Literal["file_citation"]`

4604

4605 Always `file_citation`.

4606

4607 - `"file_citation"`

4608

4609 - `end_index: Optional[int]`

4610

4611 - `file_citation: Optional[FileCitation]`

4612

4613 - `file_id: Optional[str]`

4614

4615 The ID of the specific File the citation is from.

4616

4617 - `quote: Optional[str]`

4618

4619 The specific quote in the file.

4620

4621 - `start_index: Optional[int]`

4622

4623 - `text: Optional[str]`

4624

4625 The text in the message content that needs to be replaced.

4626

4627 - `class FilePathDeltaAnnotation: …`

4628

4629 A URL for the file that's generated when the assistant used the `code_interpreter` tool to generate a file.

4630

4631 - `index: int`

4632

4633 The index of the annotation in the text content part.

4634

4635 - `type: Literal["file_path"]`

4636

4637 Always `file_path`.

4638

4639 - `"file_path"`

4640

4641 - `end_index: Optional[int]`

4642

4643 - `file_path: Optional[FilePath]`

4644

4645 - `file_id: Optional[str]`

4646

4647 The ID of the file that was generated.

4648

4649 - `start_index: Optional[int]`

4650

4651 - `text: Optional[str]`

4652

4653 The text in the message content that needs to be replaced.

4654

4655 - `value: Optional[str]`

4656

4657 The data that makes up the text.

4658

4659 - `class RefusalDeltaBlock: …`

4660

4661 The refusal content that is part of a message.

4662

4663 - `index: int`

4664

4665 The index of the refusal part in the message.

4666

4667 - `type: Literal["refusal"]`

4668

4669 Always `refusal`.

4670

4671 - `"refusal"`

4672

4673 - `refusal: Optional[str]`

4674

4675 - `class ImageURLDeltaBlock: …`

4676

4677 References an image URL in the content of a message.

4678

4679 - `index: int`

4680

4681 The index of the content part in the message.

4682

4683 - `type: Literal["image_url"]`

4684

4685 Always `image_url`.

4686

4687 - `"image_url"`

4688

4689 - `image_url: Optional[ImageURLDelta]`

4690

4691 - `detail: Optional[Literal["auto", "low", "high"]]`

4692

4693 Specifies the detail level of the image. `low` uses fewer tokens, you can opt in to high resolution using `high`.

4694

4695 - `"auto"`

4696

4697 - `"low"`

4698

4699 - `"high"`

4700

4701 - `url: Optional[str]`

4702

4703 The URL of the image, must be a supported image types: jpeg, jpg, png, gif, webp.

4704

4705 - `role: Optional[Literal["user", "assistant"]]`

4706

4707 The entity that produced the message. One of `user` or `assistant`.

4708

4709 - `"user"`

4710

4711 - `"assistant"`

4712

4713 - `object: Literal["thread.message.delta"]`

4714

4715 The object type, which is always `thread.message.delta`.

4716

4717 - `"thread.message.delta"`

4718

4719 - `event: Literal["thread.message.delta"]`

4720

4721 - `"thread.message.delta"`

4722

4723 - `class ThreadMessageCompleted: …`

4724

4725 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is completed.

4726

4727 - `data: Message`

4728

4729 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4730

4731 - `event: Literal["thread.message.completed"]`

4732

4733 - `"thread.message.completed"`

4734

4735 - `class ThreadMessageIncomplete: …`

4736

4737 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) ends before it is completed.

4738

4739 - `data: Message`

4740

4741 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

4742

4743 - `event: Literal["thread.message.incomplete"]`

4744

4745 - `"thread.message.incomplete"`

4746

4747### Run Step Stream Event

4748

4749- `RunStepStreamEvent`

4750

4751 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.

4752

4753 - `class ThreadRunStepCreated: …`

4754

4755 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.

4756

4757 - `data: RunStep`

4758

4759 Represents a step in execution of a run.

4760

4761 - `id: str`

4762

4763 The identifier of the run step, which can be referenced in API endpoints.

4764

4765 - `assistant_id: str`

4766

4767 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) associated with the run step.

4768

4769 - `cancelled_at: Optional[int]`

4770

4771 The Unix timestamp (in seconds) for when the run step was cancelled.

4772

4773 - `completed_at: Optional[int]`

4774

4775 The Unix timestamp (in seconds) for when the run step completed.

4776

4777 - `created_at: int`

4778

4779 The Unix timestamp (in seconds) for when the run step was created.

4780

4781 - `expired_at: Optional[int]`

4782

4783 The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.

4784

4785 - `failed_at: Optional[int]`

4786

4787 The Unix timestamp (in seconds) for when the run step failed.

4788

4789 - `last_error: Optional[LastError]`

4790

4791 The last error associated with this run step. Will be `null` if there are no errors.

4792

4793 - `code: Literal["server_error", "rate_limit_exceeded"]`

4794

4795 One of `server_error` or `rate_limit_exceeded`.

4796

4797 - `"server_error"`

4798

4799 - `"rate_limit_exceeded"`

4800

4801 - `message: str`

4802

4803 A human-readable description of the error.

4804

4805 - `metadata: Optional[Metadata]`

4806

4807 Set of 16 key-value pairs that can be attached to an object. This can be

4808 useful for storing additional information about the object in a structured

4809 format, and querying for objects via API or the dashboard.

4810

4811 Keys are strings with a maximum length of 64 characters. Values are strings

4812 with a maximum length of 512 characters.

4813

4814 - `object: Literal["thread.run.step"]`

4815

4816 The object type, which is always `thread.run.step`.

4817

4818 - `"thread.run.step"`

4819

4820 - `run_id: str`

4821

4822 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) that this run step is a part of.

4823

4824 - `status: Literal["in_progress", "cancelled", "failed", 2 more]`

4825

4826 The status of the run step, which can be either `in_progress`, `cancelled`, `failed`, `completed`, or `expired`.

4827

4828 - `"in_progress"`

4829

4830 - `"cancelled"`

4831

4832 - `"failed"`

4833

4834 - `"completed"`

4835

4836 - `"expired"`

4837

4838 - `step_details: StepDetails`

4839

4840 The details of the run step.

4841

4842 - `class MessageCreationStepDetails: …`

4843

4844 Details of the message creation by the run step.

4845

4846 - `message_creation: MessageCreation`

4847

4848 - `message_id: str`

4849

4850 The ID of the message that was created by this run step.

4851

4852 - `type: Literal["message_creation"]`

4853

4854 Always `message_creation`.

4855

4856 - `"message_creation"`

4857

4858 - `class ToolCallsStepDetails: …`

4859

4860 Details of the tool call.

4861

4862 - `tool_calls: List[ToolCall]`

4863

4864 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`.

4865

4866 - `class CodeInterpreterToolCall: …`

4867

4868 Details of the Code Interpreter tool call the run step was involved in.

4869

4870 - `id: str`

4871

4872 The ID of the tool call.

4873

4874 - `code_interpreter: CodeInterpreter`

4875

4876 The Code Interpreter tool call definition.

4877

4878 - `input: str`

4879

4880 The input to the Code Interpreter tool call.

4881

4882 - `outputs: List[CodeInterpreterOutput]`

4883

4884 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.

4885

4886 - `class CodeInterpreterOutputLogs: …`

4887

4888 Text output from the Code Interpreter tool call as part of a run step.

4889

4890 - `logs: str`

4891

4892 The text output from the Code Interpreter tool call.

4893

4894 - `type: Literal["logs"]`

4895

4896 Always `logs`.

4897

4898 - `"logs"`

4899

4900 - `class CodeInterpreterOutputImage: …`

4901

4902 - `image: CodeInterpreterOutputImageImage`

4903

4904 - `file_id: str`

4905

4906 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.

4907

4908 - `type: Literal["image"]`

4909

4910 Always `image`.

4911

4912 - `"image"`

4913

4914 - `type: Literal["code_interpreter"]`

4915

4916 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.

4917

4918 - `"code_interpreter"`

4919

4920 - `class FileSearchToolCall: …`

4921

4922 - `id: str`

4923

4924 The ID of the tool call object.

4925

4926 - `file_search: FileSearch`

4927

4928 For now, this is always going to be an empty object.

4929

4930 - `ranking_options: Optional[FileSearchRankingOptions]`

4931

4932 The ranking options for the file search.

4933

4934 - `ranker: Literal["auto", "default_2024_08_21"]`

4935

4936 The ranker to use for the file search. If not specified will use the `auto` ranker.

4937

4938 - `"auto"`

4939

4940 - `"default_2024_08_21"`

4941

4942 - `score_threshold: float`

4943

4944 The score threshold for the file search. All values must be a floating point number between 0 and 1.

4945

4946 - `results: Optional[List[FileSearchResult]]`

4947

4948 The results of the file search.

4949

4950 - `file_id: str`

4951

4952 The ID of the file that result was found in.

4953

4954 - `file_name: str`

4955

4956 The name of the file that result was found in.

4957

4958 - `score: float`

4959

4960 The score of the result. All values must be a floating point number between 0 and 1.

4961

4962 - `content: Optional[List[FileSearchResultContent]]`

4963

4964 The content of the result that was found. The content is only included if requested via the include query parameter.

4965

4966 - `text: Optional[str]`

4967

4968 The text content of the file.

4969

4970 - `type: Optional[Literal["text"]]`

4971

4972 The type of the content.

4973

4974 - `"text"`

4975

4976 - `type: Literal["file_search"]`

4977

4978 The type of tool call. This is always going to be `file_search` for this type of tool call.

4979

4980 - `"file_search"`

4981

4982 - `class FunctionToolCall: …`

4983

4984 - `id: str`

4985

4986 The ID of the tool call object.

4987

4988 - `function: Function`

4989

4990 The definition of the function that was called.

4991

4992 - `arguments: str`

3154 4993

3155 The delta containing the fields that have changed on the Message.4994 The arguments passed to the function.

3156 4995

3157 - `content: Optional[List[object]]`4996 - `name: str`

3158 4997

3159 The content of the message in array of text and/or images.4998 The name of the function.

3160 4999

3161 - `role: Optional[Literal["user", "assistant"]]`5000 - `output: Optional[str]`

3162 5001

3163 The entity that produced the message. One of `user` or `assistant`.5002 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.

3164 5003

3165 - `"user"`5004 - `type: Literal["function"]`

3166 5005

3167 - `"assistant"`5006 The type of tool call. This is always going to be `function` for this type of tool call.

3168 5007

3169 - `object: Literal["thread.message.delta"]`5008 - `"function"`

3170 5009

3171 The object type, which is always `thread.message.delta`.5010 - `type: Literal["tool_calls"]`

3172 5011

3173 - `"thread.message.delta"`5012 Always `tool_calls`.

3174 5013

3175 - `event: Literal["thread.message.delta"]`5014 - `"tool_calls"`

3176 5015

3177 - `"thread.message.delta"`5016 - `thread_id: str`

3178 5017

3179 - `class ThreadMessageCompleted: …`5018 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.

3180 5019

3181 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is completed.5020 - `type: Literal["message_creation", "tool_calls"]`

3182 5021

3183 - `data: Message`5022 The type of run step, which can be either `message_creation` or `tool_calls`.

3184 5023

3185 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).5024 - `"message_creation"`

3186 5025

3187 - `event: Literal["thread.message.completed"]`5026 - `"tool_calls"`

3188 5027

3189 - `"thread.message.completed"`5028 - `usage: Optional[Usage]`

3190 5029

3191 - `class ThreadMessageIncomplete: …`5030 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.

3192 5031

3193 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) ends before it is completed.5032 - `completion_tokens: int`

3194 5033

3195 - `data: Message`5034 Number of completion tokens used over the course of the run step.

3196 5035

3197 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).5036 - `prompt_tokens: int`

3198 5037

3199 - `event: Literal["thread.message.incomplete"]`5038 Number of prompt tokens used over the course of the run step.

3200 5039

3201 - `"thread.message.incomplete"`5040 - `total_tokens: int`

3202 5041

3203### Run Step Stream Event5042 Total number of tokens used (prompt + completion).

3204 5043

3205- `RunStepStreamEvent`5044 - `event: Literal["thread.run.step.created"]`

3206 5045

3207 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.5046 - `"thread.run.step.created"`

3208 5047

3209 - `class ThreadRunStepCreated: …`5048 - `class ThreadRunStepInProgress: …`

3210 5049

3211 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.5050 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) moves to an `in_progress` state.

3212 5051

3213 - `data: RunStep`5052 - `data: RunStep`

3214 5053

3215 Represents a step in execution of a run.5054 Represents a step in execution of a run.

3216 5055

3217 - `id: str`5056 - `event: Literal["thread.run.step.in_progress"]`

3218 5057

3219 The identifier of the run step, which can be referenced in API endpoints.5058 - `"thread.run.step.in_progress"`

3220 5059

3221 - `assistant_id: str`5060 - `class ThreadRunStepDelta: …`

3222 5061

3223 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) associated with the run step.5062 Occurs when parts of a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) are being streamed.

3224 5063

3225 - `cancelled_at: Optional[int]`5064 - `data: RunStepDeltaEvent`

3226 5065

3227 The Unix timestamp (in seconds) for when the run step was cancelled.5066 Represents a run step delta i.e. any changed fields on a run step during streaming.

3228 5067

3229 - `completed_at: Optional[int]`5068 - `id: str`

3230 5069

3231 The Unix timestamp (in seconds) for when the run step completed.5070 The identifier of the run step, which can be referenced in API endpoints.

3232 5071

3233 - `created_at: int`5072 - `delta: RunStepDelta`

3234 5073

3235 The Unix timestamp (in seconds) for when the run step was created.5074 The delta containing the fields that have changed on the run step.

3236 5075

3237 - `expired_at: Optional[int]`5076 - `step_details: Optional[StepDetails]`

3238 5077

3239 The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.5078 The details of the run step.

3240 5079

3241 - `failed_at: Optional[int]`5080 - `class RunStepDeltaMessageDelta: …`

3242 5081

3243 The Unix timestamp (in seconds) for when the run step failed.5082 Details of the message creation by the run step.

3244 5083

3245 - `last_error: Optional[LastError]`5084 - `type: Literal["message_creation"]`

3246 5085

3247 The last error associated with this run step. Will be `null` if there are no errors.5086 Always `message_creation`.

3248 5087

3249 - `code: Literal["server_error", "rate_limit_exceeded"]`5088 - `"message_creation"`

3250 5089

3251 One of `server_error` or `rate_limit_exceeded`.5090 - `message_creation: Optional[MessageCreation]`

3252 5091

3253 - `"server_error"`5092 - `message_id: Optional[str]`

3254 5093

3255 - `"rate_limit_exceeded"`5094 The ID of the message that was created by this run step.

3256 5095

3257 - `message: str`5096 - `class ToolCallDeltaObject: …`

3258 5097

3259 A human-readable description of the error.5098 Details of the tool call.

3260 5099

3261 - `metadata: Optional[Metadata]`5100 - `type: Literal["tool_calls"]`

3262 5101

3263 Set of 16 key-value pairs that can be attached to an object. This can be5102 Always `tool_calls`.

3264 useful for storing additional information about the object in a structured

3265 format, and querying for objects via API or the dashboard.

3266 5103

3267 Keys are strings with a maximum length of 64 characters. Values are strings5104 - `"tool_calls"`

3268 with a maximum length of 512 characters.

3269 5105

3270 - `object: Literal["thread.run.step"]`5106 - `tool_calls: Optional[List[ToolCallDelta]]`

3271 5107

3272 The object type, which is always `thread.run.step`.5108 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`.

3273 5109

3274 - `"thread.run.step"`5110 - `class CodeInterpreterToolCallDelta: …`

3275 5111

3276 - `run_id: str`5112 Details of the Code Interpreter tool call the run step was involved in.

3277 5113

3278 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) that this run step is a part of.5114 - `index: int`

3279 5115

3280 - `status: Literal["in_progress", "cancelled", "failed", 2 more]`5116 The index of the tool call in the tool calls array.

3281 5117

3282 The status of the run step, which can be either `in_progress`, `cancelled`, `failed`, `completed`, or `expired`.5118 - `type: Literal["code_interpreter"]`

3283 5119

3284 - `"in_progress"`5120 The type of tool call. This is always going to be `code_interpreter` for this type of tool call.

3285 5121

3286 - `"cancelled"`5122 - `"code_interpreter"`

3287 5123

3288 - `"failed"`5124 - `id: Optional[str]`

3289 5125

3290 - `"completed"`5126 The ID of the tool call.

3291 5127

3292 - `"expired"`5128 - `code_interpreter: Optional[CodeInterpreter]`

3293 5129

3294 - `step_details: StepDetails`5130 The Code Interpreter tool call definition.

3295 5131

3296 The details of the run step.5132 - `input: Optional[str]`

3297 5133

3298 - `class MessageCreationStepDetails: …`5134 The input to the Code Interpreter tool call.

3299 5135

3300 Details of the message creation by the run step.5136 - `outputs: Optional[List[CodeInterpreterOutput]]`

3301 5137

3302 - `message_creation: MessageCreation`5138 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.

3303 5139

3304 - `message_id: str`5140 - `class CodeInterpreterLogs: …`

3305 5141

3306 The ID of the message that was created by this run step.5142 Text output from the Code Interpreter tool call as part of a run step.

3307 5143

3308 - `type: Literal["message_creation"]`5144 - `index: int`

3309 5145

3310 Always `message_creation`.5146 The index of the output in the outputs array.

3311 5147

3312 - `"message_creation"`5148 - `type: Literal["logs"]`

3313 5149

3314 - `class ToolCallsStepDetails: …`5150 Always `logs`.

3315 5151

3316 Details of the tool call.5152 - `"logs"`

3317 5153

3318 - `tool_calls: List[object]`5154 - `logs: Optional[str]`

3319 5155

3320 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`.5156 The text output from the Code Interpreter tool call.

3321 5157

3322 - `type: Literal["tool_calls"]`5158 - `class CodeInterpreterOutputImage: …`

3323 5159

3324 Always `tool_calls`.5160 - `index: int`

3325 5161

3326 - `"tool_calls"`5162 The index of the output in the outputs array.

3327 5163

3328 - `thread_id: str`5164 - `type: Literal["image"]`

3329 5165

3330 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.5166 Always `image`.

3331 5167

3332 - `type: Literal["message_creation", "tool_calls"]`5168 - `"image"`

3333 5169

3334 The type of run step, which can be either `message_creation` or `tool_calls`.5170 - `image: Optional[Image]`

3335 5171

3336 - `"message_creation"`5172 - `file_id: Optional[str]`

3337 5173

3338 - `"tool_calls"`5174 The [file](https://platform.openai.com/docs/api-reference/files) ID of the image.

3339 5175

3340 - `usage: Optional[Usage]`5176 - `class FileSearchToolCallDelta: …`

3341 5177

3342 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.5178 - `file_search: object`

3343 5179

3344 - `completion_tokens: int`5180 For now, this is always going to be an empty object.

3345 5181

3346 Number of completion tokens used over the course of the run step.5182 - `index: int`

3347 5183

3348 - `prompt_tokens: int`5184 The index of the tool call in the tool calls array.

3349 5185

3350 Number of prompt tokens used over the course of the run step.5186 - `type: Literal["file_search"]`

3351 5187

3352 - `total_tokens: int`5188 The type of tool call. This is always going to be `file_search` for this type of tool call.

3353 5189

3354 Total number of tokens used (prompt + completion).5190 - `"file_search"`

3355 5191

3356 - `event: Literal["thread.run.step.created"]`5192 - `id: Optional[str]`

3357 5193

3358 - `"thread.run.step.created"`5194 The ID of the tool call object.

3359 5195

3360 - `class ThreadRunStepInProgress: …`5196 - `class FunctionToolCallDelta: …`

3361 5197

3362 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) moves to an `in_progress` state.5198 - `index: int`

3363 5199

3364 - `data: RunStep`5200 The index of the tool call in the tool calls array.

3365 5201

3366 Represents a step in execution of a run.5202 - `type: Literal["function"]`

3367 5203

3368 - `event: Literal["thread.run.step.in_progress"]`5204 The type of tool call. This is always going to be `function` for this type of tool call.

3369 5205

3370 - `"thread.run.step.in_progress"`5206 - `"function"`

3371 5207

3372 - `class ThreadRunStepDelta: …`5208 - `id: Optional[str]`

3373 5209

3374 Occurs when parts of a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) are being streamed.5210 The ID of the tool call object.

3375 5211

3376 - `data: RunStepDeltaEvent`5212 - `function: Optional[Function]`

3377 5213

3378 Represents a run step delta i.e. any changed fields on a run step during streaming.5214 The definition of the function that was called.

3379 5215

3380 - `id: str`5216 - `arguments: Optional[str]`

3381 5217

3382 The identifier of the run step, which can be referenced in API endpoints.5218 The arguments passed to the function.

5219

5220 - `name: Optional[str]`

5221

5222 The name of the function.

5223

5224 - `output: Optional[str]`

3383 5225

3384 - `delta: object`5226 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.

3385 5227

3386 - `object: Literal["thread.run.step.delta"]`5228 - `object: Literal["thread.run.step.delta"]`

3387 5229

3669 5511

3670 The Unix timestamp (in seconds) for when the run was started.5512 The Unix timestamp (in seconds) for when the run was started.

3671 5513

3672 - `status: object`5514 - `status: RunStatus`

5515

5516 The status of the run, which can be either `queued`, `in_progress`, `requires_action`, `cancelling`, `cancelled`, `failed`, `completed`, `incomplete`, or `expired`.

5517

5518 - `"queued"`

5519

5520 - `"in_progress"`

5521

5522 - `"requires_action"`

5523

5524 - `"cancelling"`

5525

5526 - `"cancelled"`

5527

5528 - `"failed"`

5529

5530 - `"completed"`

5531

5532 - `"incomplete"`

5533

5534 - `"expired"`

3673 5535

3674 - `thread_id: str`5536 - `thread_id: str`

3675 5537

3713 5575

3714 The name of the function to call.5576 The name of the function to call.

3715 5577

3716 - `tools: List[object]`5578 - `tools: List[AssistantTool]`

3717 5579

3718 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.5580 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

3719 5581

5582 - `class CodeInterpreterTool: …`

5583

5584 - `type: Literal["code_interpreter"]`

5585

5586 The type of tool being defined: `code_interpreter`

5587

5588 - `"code_interpreter"`

5589

5590 - `class FileSearchTool: …`

5591

5592 - `type: Literal["file_search"]`

5593

5594 The type of tool being defined: `file_search`

5595

5596 - `"file_search"`

5597

5598 - `file_search: Optional[FileSearch]`

5599

5600 Overrides for the file search tool.

5601

5602 - `max_num_results: Optional[int]`

5603

5604 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.

5605

5606 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.

5607

5608 - `ranking_options: Optional[FileSearchRankingOptions]`

5609

5610 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.

5611

5612 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

5613

5614 - `score_threshold: float`

5615

5616 The score threshold for the file search. All values must be a floating point number between 0 and 1.

5617

5618 - `ranker: Optional[Literal["auto", "default_2024_08_21"]]`

5619

5620 The ranker to use for the file search. If not specified will use the `auto` ranker.

5621

5622 - `"auto"`

5623

5624 - `"default_2024_08_21"`

5625

5626 - `class FunctionTool: …`

5627

5628 - `function: FunctionDefinition`

5629

5630 - `name: str`

5631

5632 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.

5633

5634 - `description: Optional[str]`

5635

5636 A description of what the function does, used by the model to choose when and how to call the function.

5637

5638 - `parameters: Optional[FunctionParameters]`

5639

5640 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.

5641

5642 Omitting `parameters` defines a function with an empty parameter list.

5643

5644 - `strict: Optional[bool]`

5645

5646 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).

5647

5648 - `type: Literal["function"]`

5649

5650 The type of tool being defined: `function`

5651

5652 - `"function"`

5653

3720 - `truncation_strategy: Optional[TruncationStrategy]`5654 - `truncation_strategy: Optional[TruncationStrategy]`

3721 5655

3722 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.5656 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

python/resources/beta/subresources/assistants/index.md 2026-07-10 23:02 UTC to 2026-07-12 06:58 UTC

Assistants

List assistants

Parameters

Returns

Example

Response

Example

Response

Create assistant

Parameters

Returns

Example

Response

Code Interpreter

Response

Files

Response

Retrieve assistant

Parameters

Returns

Example

Response

Example

Response

Modify assistant

Parameters

Returns

Example

Response

Example

Response

Delete assistant

Parameters

Returns

Example

Response

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