Assistants

List assistants

client.Beta.Assistants.List(ctx, query) (*CursorPage[Assistant], error)

get /assistants

List assistants

Parameters

query BetaAssistantListParams
- After param.Field[string]
  
  A cursor for use in pagination. after is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, ending with obj_foo, your subsequent call can include after=obj_foo in order to fetch the next page of the list.
- Before param.Field[string]
  
  A cursor for use in pagination. before is an object ID that defines your place in the list. For instance, if you make a list request and receive 100 objects, starting with obj_foo, your subsequent call can include before=obj_foo in order to fetch the previous page of the list.
- Limit param.Field[int64]
  
  A limit on the number of objects to be returned. Limit can range between 1 and 100, and the default is 20.
- Order param.Field[BetaAssistantListParamsOrder]
  
  Sort order by the created_at timestamp of the objects. asc for ascending order and desc for descending order.
  - const BetaAssistantListParamsOrderAsc BetaAssistantListParamsOrder = "asc"
  - const BetaAssistantListParamsOrderDesc BetaAssistantListParamsOrder = "desc"

Returns

type Assistant struct{…}

Represents an assistant that can call the model and use tools.
- ID string
  
  The identifier, which can be referenced in API endpoints.
- CreatedAt int64
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Description string
  
  The description of the assistant. The maximum length is 512 characters.
- Instructions string
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Metadata Metadata
  
  Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
  
  Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
- Model string
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- Name string
  
  The name of the assistant. The maximum length is 256 characters.
- Object Assistant
  
  The object type, which is always assistant.
  - const AssistantAssistant Assistant = "assistant"
- Tools []any
  
  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.
- ResponseFormat AssistantResponseFormatOptionUnion
  
  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.
  - type Auto string
    
    auto is the default value
    - const AutoAuto Auto = "auto"
  - type ResponseFormatText struct{…}
    
    Default response format. Used to generate text responses.
    - Type Text
      
      The type of response format being defined. Always text.
      - const TextText Text = "text"
  - type ResponseFormatJSONObject struct{…}
    
    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 JSONObject
      
      The type of response format being defined. Always json_object.
      - const JSONObjectJSONObject JSONObject = "json_object"
  - type ResponseFormatJSONSchema struct{…}
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JSONSchema ResponseFormatJSONSchemaJSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - Name string
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - Description string
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Schema map[string, any]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Strict 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 JSONSchema
      
      The type of response format being defined. Always json_schema.
      - const JSONSchemaJSONSchema JSONSchema = "json_schema"
- Temperature float64
  
  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.
- ToolResources AssistantToolResources
  
  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.
  - CodeInterpreter AssistantToolResourcesCodeInterpreter
    - FileIDs []string
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - FileSearch AssistantToolResourcesFileSearch
    - VectorStoreIDs []string
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- TopP float64
  
  An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
  
  We generally recommend altering this or temperature but not both.

Example

package main

import (
  "context"
  "fmt"

  "github.com/openai/openai-go"
  "github.com/openai/openai-go/option"
)

func main() {
  client := openai.NewClient(
    option.WithAPIKey("My API Key"),
  )
  page, err := client.Beta.Assistants.List(context.TODO(), openai.BetaAssistantListParams{

  })
  if err != nil {
    panic(err.Error())
  }
  fmt.Printf("%+v\n", page)
}

Response

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

Create assistant

client.Beta.Assistants.New(ctx, body) (*Assistant, error)

post /assistants

Create assistant

Parameters

body BetaAssistantNewParams
- Model param.Field[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.
  - string
  - type ChatModel string
    - const ChatModelGPT5_6Sol ChatModel = "gpt-5.6-sol"
    - const ChatModelGPT5_6Terra ChatModel = "gpt-5.6-terra"
    - const ChatModelGPT5_6Luna ChatModel = "gpt-5.6-luna"
    - const ChatModelGPT5_4 ChatModel = "gpt-5.4"
    - const ChatModelGPT5_4Mini ChatModel = "gpt-5.4-mini"
    - const ChatModelGPT5_4Nano ChatModel = "gpt-5.4-nano"
    - const ChatModelGPT5_4Mini2026_03_17 ChatModel = "gpt-5.4-mini-2026-03-17"
    - const ChatModelGPT5_4Nano2026_03_17 ChatModel = "gpt-5.4-nano-2026-03-17"
    - const ChatModelGPT5_3ChatLatest ChatModel = "gpt-5.3-chat-latest"
    - const ChatModelGPT5_2 ChatModel = "gpt-5.2"
    - const ChatModelGPT5_2_2025_12_11 ChatModel = "gpt-5.2-2025-12-11"
    - const ChatModelGPT5_2ChatLatest ChatModel = "gpt-5.2-chat-latest"
    - const ChatModelGPT5_2Pro ChatModel = "gpt-5.2-pro"
    - const ChatModelGPT5_2Pro2025_12_11 ChatModel = "gpt-5.2-pro-2025-12-11"
    - const ChatModelGPT5_1 ChatModel = "gpt-5.1"
    - const ChatModelGPT5_1_2025_11_13 ChatModel = "gpt-5.1-2025-11-13"
    - const ChatModelGPT5_1Codex ChatModel = "gpt-5.1-codex"
    - const ChatModelGPT5_1Mini ChatModel = "gpt-5.1-mini"
    - const ChatModelGPT5_1ChatLatest ChatModel = "gpt-5.1-chat-latest"
    - const ChatModelGPT5 ChatModel = "gpt-5"
    - const ChatModelGPT5Mini ChatModel = "gpt-5-mini"
    - const ChatModelGPT5Nano ChatModel = "gpt-5-nano"
    - const ChatModelGPT5_2025_08_07 ChatModel = "gpt-5-2025-08-07"
    - const ChatModelGPT5Mini2025_08_07 ChatModel = "gpt-5-mini-2025-08-07"
    - const ChatModelGPT5Nano2025_08_07 ChatModel = "gpt-5-nano-2025-08-07"
    - const ChatModelGPT5ChatLatest ChatModel = "gpt-5-chat-latest"
    - const ChatModelGPT4_1 ChatModel = "gpt-4.1"
    - const ChatModelGPT4_1Mini ChatModel = "gpt-4.1-mini"
    - const ChatModelGPT4_1Nano ChatModel = "gpt-4.1-nano"
    - const ChatModelGPT4_1_2025_04_14 ChatModel = "gpt-4.1-2025-04-14"
    - const ChatModelGPT4_1Mini2025_04_14 ChatModel = "gpt-4.1-mini-2025-04-14"
    - const ChatModelGPT4_1Nano2025_04_14 ChatModel = "gpt-4.1-nano-2025-04-14"
    - const ChatModelO4Mini ChatModel = "o4-mini"
    - const ChatModelO4Mini2025_04_16 ChatModel = "o4-mini-2025-04-16"
    - const ChatModelO3 ChatModel = "o3"
    - const ChatModelO3_2025_04_16 ChatModel = "o3-2025-04-16"
    - const ChatModelO3Mini ChatModel = "o3-mini"
    - const ChatModelO3Mini2025_01_31 ChatModel = "o3-mini-2025-01-31"
    - const ChatModelO1 ChatModel = "o1"
    - const ChatModelO1_2024_12_17 ChatModel = "o1-2024-12-17"
    - const ChatModelO1Preview ChatModel = "o1-preview"
    - const ChatModelO1Preview2024_09_12 ChatModel = "o1-preview-2024-09-12"
    - const ChatModelO1Mini ChatModel = "o1-mini"
    - const ChatModelO1Mini2024_09_12 ChatModel = "o1-mini-2024-09-12"
    - const ChatModelGPT4o ChatModel = "gpt-4o"
    - const ChatModelGPT4o2024_11_20 ChatModel = "gpt-4o-2024-11-20"
    - const ChatModelGPT4o2024_08_06 ChatModel = "gpt-4o-2024-08-06"
    - const ChatModelGPT4o2024_05_13 ChatModel = "gpt-4o-2024-05-13"
    - const ChatModelGPT4oAudioPreview ChatModel = "gpt-4o-audio-preview"
    - const ChatModelGPT4oAudioPreview2024_10_01 ChatModel = "gpt-4o-audio-preview-2024-10-01"
    - const ChatModelGPT4oAudioPreview2024_12_17 ChatModel = "gpt-4o-audio-preview-2024-12-17"
    - const ChatModelGPT4oAudioPreview2025_06_03 ChatModel = "gpt-4o-audio-preview-2025-06-03"
    - const ChatModelGPT4oMiniAudioPreview ChatModel = "gpt-4o-mini-audio-preview"
    - const ChatModelGPT4oMiniAudioPreview2024_12_17 ChatModel = "gpt-4o-mini-audio-preview-2024-12-17"
    - const ChatModelGPT4oSearchPreview ChatModel = "gpt-4o-search-preview"
    - const ChatModelGPT4oMiniSearchPreview ChatModel = "gpt-4o-mini-search-preview"
    - const ChatModelGPT4oSearchPreview2025_03_11 ChatModel = "gpt-4o-search-preview-2025-03-11"
    - const ChatModelGPT4oMiniSearchPreview2025_03_11 ChatModel = "gpt-4o-mini-search-preview-2025-03-11"
    - const ChatModelChatgpt4oLatest ChatModel = "chatgpt-4o-latest"
    - const ChatModelCodexMiniLatest ChatModel = "codex-mini-latest"
    - const ChatModelGPT4oMini ChatModel = "gpt-4o-mini"
    - const ChatModelGPT4oMini2024_07_18 ChatModel = "gpt-4o-mini-2024-07-18"
    - const ChatModelGPT4Turbo ChatModel = "gpt-4-turbo"
    - const ChatModelGPT4Turbo2024_04_09 ChatModel = "gpt-4-turbo-2024-04-09"
    - const ChatModelGPT4_0125Preview ChatModel = "gpt-4-0125-preview"
    - const ChatModelGPT4TurboPreview ChatModel = "gpt-4-turbo-preview"
    - const ChatModelGPT4_1106Preview ChatModel = "gpt-4-1106-preview"
    - const ChatModelGPT4VisionPreview ChatModel = "gpt-4-vision-preview"
    - const ChatModelGPT4 ChatModel = "gpt-4"
    - const ChatModelGPT4_0314 ChatModel = "gpt-4-0314"
    - const ChatModelGPT4_0613 ChatModel = "gpt-4-0613"
    - const ChatModelGPT4_32k ChatModel = "gpt-4-32k"
    - const ChatModelGPT4_32k0314 ChatModel = "gpt-4-32k-0314"
    - const ChatModelGPT4_32k0613 ChatModel = "gpt-4-32k-0613"
    - const ChatModelGPT3_5Turbo ChatModel = "gpt-3.5-turbo"
    - const ChatModelGPT3_5Turbo16k ChatModel = "gpt-3.5-turbo-16k"
    - const ChatModelGPT3_5Turbo0301 ChatModel = "gpt-3.5-turbo-0301"
    - const ChatModelGPT3_5Turbo0613 ChatModel = "gpt-3.5-turbo-0613"
    - const ChatModelGPT3_5Turbo1106 ChatModel = "gpt-3.5-turbo-1106"
    - const ChatModelGPT3_5Turbo0125 ChatModel = "gpt-3.5-turbo-0125"
    - const ChatModelGPT3_5Turbo16k0613 ChatModel = "gpt-3.5-turbo-16k-0613"
- Description param.Field[string]
  
  The description of the assistant. The maximum length is 512 characters.
- Instructions param.Field[string]
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Metadata param.Field[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 param.Field[string]
  
  The name of the assistant. The maximum length is 256 characters.
- ReasoningEffort param.Field[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.
- ResponseFormat param.Field[AssistantResponseFormatOptionUnion]
  
  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.
- Temperature param.Field[float64]
  
  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.
- ToolResources param.Field[BetaAssistantNewParamsToolResources]
  
  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.
  - CodeInterpreter BetaAssistantNewParamsToolResourcesCodeInterpreter
    - FileIDs []string
      
      A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
  - FileSearch BetaAssistantNewParamsToolResourcesFileSearch
    - VectorStoreIDs []string
      
      The vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
    - VectorStores []BetaAssistantNewParamsToolResourcesFileSearchVectorStore
      
      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.
      - ChunkingStrategy BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyUnion
        
        The chunking strategy used to chunk the file(s). If not set, will use the auto strategy.
        
        type BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyAuto struct{…}
        
        The default strategy. This strategy currently uses a max_chunk_size_tokens of 800 and chunk_overlap_tokens of 400.
        
        Type Auto
        
        Always auto.
        
        const AutoAuto Auto = "auto"
        
        type BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyStatic struct{…}
        
        Static BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyStaticStatic
        
        ChunkOverlapTokens int64
        
        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.
        
        MaxChunkSizeTokens int64
        
        The maximum number of tokens in each chunk. The default value is 800. The minimum value is 100 and the maximum value is 4096.
        
        Type Static
        
        Always static.
        
        const StaticStatic Static = "static"
      - FileIDs []string
        
        A list of file IDs to add to the vector store. For vector stores created before Nov 2025, there can be a maximum of 10,000 files in a vector store. For vector stores created starting in Nov 2025, the limit is 100,000,000 files.
      - Metadata 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 param.Field[[]any]
  
  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.
- TopP param.Field[float64]
  
  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

type Assistant struct{…}

Represents an assistant that can call the model and use tools.
- ID string
  
  The identifier, which can be referenced in API endpoints.
- CreatedAt int64
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Description string
  
  The description of the assistant. The maximum length is 512 characters.
- Instructions string
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Metadata Metadata
  
  Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
  
  Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
- Model string
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- Name string
  
  The name of the assistant. The maximum length is 256 characters.
- Object Assistant
  
  The object type, which is always assistant.
  - const AssistantAssistant Assistant = "assistant"
- Tools []any
  
  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.
- ResponseFormat AssistantResponseFormatOptionUnion
  
  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.
  - type Auto string
    
    auto is the default value
    - const AutoAuto Auto = "auto"
  - type ResponseFormatText struct{…}
    
    Default response format. Used to generate text responses.
    - Type Text
      
      The type of response format being defined. Always text.
      - const TextText Text = "text"
  - type ResponseFormatJSONObject struct{…}
    
    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 JSONObject
      
      The type of response format being defined. Always json_object.
      - const JSONObjectJSONObject JSONObject = "json_object"
  - type ResponseFormatJSONSchema struct{…}
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JSONSchema ResponseFormatJSONSchemaJSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - Name string
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - Description string
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Schema map[string, any]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Strict 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 JSONSchema
      
      The type of response format being defined. Always json_schema.
      - const JSONSchemaJSONSchema JSONSchema = "json_schema"
- Temperature float64
  
  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.
- ToolResources AssistantToolResources
  
  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.
  - CodeInterpreter AssistantToolResourcesCodeInterpreter
    - FileIDs []string
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - FileSearch AssistantToolResourcesFileSearch
    - VectorStoreIDs []string
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- TopP float64
  
  An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
  
  We generally recommend altering this or temperature but not both.

Example

package main

import (
  "context"
  "fmt"

  "github.com/openai/openai-go"
  "github.com/openai/openai-go/option"
  "github.com/openai/openai-go/shared"
)

func main() {
  client := openai.NewClient(
    option.WithAPIKey("My API Key"),
  )
  assistant, err := client.Beta.Assistants.New(context.TODO(), openai.BetaAssistantNewParams{
    Model: shared.ChatModelGPT4o,
  })
  if err != nil {
    panic(err.Error())
  }
  fmt.Printf("%+v\n", assistant.ID)
}

Response

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

Retrieve assistant

client.Beta.Assistants.Get(ctx, assistantID) (*Assistant, error)

get /assistants/{assistant_id}

Retrieve assistant

Parameters

assistantID string

Returns

type Assistant struct{…}

Represents an assistant that can call the model and use tools.
- ID string
  
  The identifier, which can be referenced in API endpoints.
- CreatedAt int64
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Description string
  
  The description of the assistant. The maximum length is 512 characters.
- Instructions string
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Metadata Metadata
  
  Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
  
  Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
- Model string
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- Name string
  
  The name of the assistant. The maximum length is 256 characters.
- Object Assistant
  
  The object type, which is always assistant.
  - const AssistantAssistant Assistant = "assistant"
- Tools []any
  
  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.
- ResponseFormat AssistantResponseFormatOptionUnion
  
  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.
  - type Auto string
    
    auto is the default value
    - const AutoAuto Auto = "auto"
  - type ResponseFormatText struct{…}
    
    Default response format. Used to generate text responses.
    - Type Text
      
      The type of response format being defined. Always text.
      - const TextText Text = "text"
  - type ResponseFormatJSONObject struct{…}
    
    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 JSONObject
      
      The type of response format being defined. Always json_object.
      - const JSONObjectJSONObject JSONObject = "json_object"
  - type ResponseFormatJSONSchema struct{…}
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JSONSchema ResponseFormatJSONSchemaJSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - Name string
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - Description string
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Schema map[string, any]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Strict 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 JSONSchema
      
      The type of response format being defined. Always json_schema.
      - const JSONSchemaJSONSchema JSONSchema = "json_schema"
- Temperature float64
  
  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.
- ToolResources AssistantToolResources
  
  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.
  - CodeInterpreter AssistantToolResourcesCodeInterpreter
    - FileIDs []string
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - FileSearch AssistantToolResourcesFileSearch
    - VectorStoreIDs []string
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- TopP float64
  
  An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
  
  We generally recommend altering this or temperature but not both.

Example

package main

import (
  "context"
  "fmt"

  "github.com/openai/openai-go"
  "github.com/openai/openai-go/option"
)

func main() {
  client := openai.NewClient(
    option.WithAPIKey("My API Key"),
  )
  assistant, err := client.Beta.Assistants.Get(context.TODO(), "assistant_id")
  if err != nil {
    panic(err.Error())
  }
  fmt.Printf("%+v\n", assistant.ID)
}

Response

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

Modify assistant

client.Beta.Assistants.Update(ctx, assistantID, body) (*Assistant, error)

post /assistants/{assistant_id}

Modify assistant

Parameters

assistantID string
body BetaAssistantUpdateParams
- Description param.Field[string]
  
  The description of the assistant. The maximum length is 512 characters.
- Instructions param.Field[string]
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Metadata param.Field[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 param.Field[BetaAssistantUpdateParamsModel]
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
  - string
  - type BetaAssistantUpdateParamsModel string
    
    ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
    - const BetaAssistantUpdateParamsModelGPT5 BetaAssistantUpdateParamsModel = "gpt-5"
    - const BetaAssistantUpdateParamsModelGPT5Mini BetaAssistantUpdateParamsModel = "gpt-5-mini"
    - const BetaAssistantUpdateParamsModelGPT5Nano BetaAssistantUpdateParamsModel = "gpt-5-nano"
    - const BetaAssistantUpdateParamsModelGPT5_2025_08_07 BetaAssistantUpdateParamsModel = "gpt-5-2025-08-07"
    - const BetaAssistantUpdateParamsModelGPT5Mini2025_08_07 BetaAssistantUpdateParamsModel = "gpt-5-mini-2025-08-07"
    - const BetaAssistantUpdateParamsModelGPT5Nano2025_08_07 BetaAssistantUpdateParamsModel = "gpt-5-nano-2025-08-07"
    - const BetaAssistantUpdateParamsModelGPT4_1 BetaAssistantUpdateParamsModel = "gpt-4.1"
    - const BetaAssistantUpdateParamsModelGPT4_1Mini BetaAssistantUpdateParamsModel = "gpt-4.1-mini"
    - const BetaAssistantUpdateParamsModelGPT4_1Nano BetaAssistantUpdateParamsModel = "gpt-4.1-nano"
    - const BetaAssistantUpdateParamsModelGPT4_1_2025_04_14 BetaAssistantUpdateParamsModel = "gpt-4.1-2025-04-14"
    - const BetaAssistantUpdateParamsModelGPT4_1Mini2025_04_14 BetaAssistantUpdateParamsModel = "gpt-4.1-mini-2025-04-14"
    - const BetaAssistantUpdateParamsModelGPT4_1Nano2025_04_14 BetaAssistantUpdateParamsModel = "gpt-4.1-nano-2025-04-14"
    - const BetaAssistantUpdateParamsModelO3Mini BetaAssistantUpdateParamsModel = "o3-mini"
    - const BetaAssistantUpdateParamsModelO3Mini2025_01_31 BetaAssistantUpdateParamsModel = "o3-mini-2025-01-31"
    - const BetaAssistantUpdateParamsModelO1 BetaAssistantUpdateParamsModel = "o1"
    - const BetaAssistantUpdateParamsModelO1_2024_12_17 BetaAssistantUpdateParamsModel = "o1-2024-12-17"
    - const BetaAssistantUpdateParamsModelGPT4o BetaAssistantUpdateParamsModel = "gpt-4o"
    - const BetaAssistantUpdateParamsModelGPT4o2024_11_20 BetaAssistantUpdateParamsModel = "gpt-4o-2024-11-20"
    - const BetaAssistantUpdateParamsModelGPT4o2024_08_06 BetaAssistantUpdateParamsModel = "gpt-4o-2024-08-06"
    - const BetaAssistantUpdateParamsModelGPT4o2024_05_13 BetaAssistantUpdateParamsModel = "gpt-4o-2024-05-13"
    - const BetaAssistantUpdateParamsModelGPT4oMini BetaAssistantUpdateParamsModel = "gpt-4o-mini"
    - const BetaAssistantUpdateParamsModelGPT4oMini2024_07_18 BetaAssistantUpdateParamsModel = "gpt-4o-mini-2024-07-18"
    - const BetaAssistantUpdateParamsModelGPT4_5Preview BetaAssistantUpdateParamsModel = "gpt-4.5-preview"
    - const BetaAssistantUpdateParamsModelGPT4_5Preview2025_02_27 BetaAssistantUpdateParamsModel = "gpt-4.5-preview-2025-02-27"
    - const BetaAssistantUpdateParamsModelGPT4Turbo BetaAssistantUpdateParamsModel = "gpt-4-turbo"
    - const BetaAssistantUpdateParamsModelGPT4Turbo2024_04_09 BetaAssistantUpdateParamsModel = "gpt-4-turbo-2024-04-09"
    - const BetaAssistantUpdateParamsModelGPT4_0125Preview BetaAssistantUpdateParamsModel = "gpt-4-0125-preview"
    - const BetaAssistantUpdateParamsModelGPT4TurboPreview BetaAssistantUpdateParamsModel = "gpt-4-turbo-preview"
    - const BetaAssistantUpdateParamsModelGPT4_1106Preview BetaAssistantUpdateParamsModel = "gpt-4-1106-preview"
    - const BetaAssistantUpdateParamsModelGPT4VisionPreview BetaAssistantUpdateParamsModel = "gpt-4-vision-preview"
    - const BetaAssistantUpdateParamsModelGPT4 BetaAssistantUpdateParamsModel = "gpt-4"
    - const BetaAssistantUpdateParamsModelGPT4_0314 BetaAssistantUpdateParamsModel = "gpt-4-0314"
    - const BetaAssistantUpdateParamsModelGPT4_0613 BetaAssistantUpdateParamsModel = "gpt-4-0613"
    - const BetaAssistantUpdateParamsModelGPT4_32k BetaAssistantUpdateParamsModel = "gpt-4-32k"
    - const BetaAssistantUpdateParamsModelGPT4_32k0314 BetaAssistantUpdateParamsModel = "gpt-4-32k-0314"
    - const BetaAssistantUpdateParamsModelGPT4_32k0613 BetaAssistantUpdateParamsModel = "gpt-4-32k-0613"
    - const BetaAssistantUpdateParamsModelGPT3_5Turbo BetaAssistantUpdateParamsModel = "gpt-3.5-turbo"
    - const BetaAssistantUpdateParamsModelGPT3_5Turbo16k BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-16k"
    - const BetaAssistantUpdateParamsModelGPT3_5Turbo0613 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-0613"
    - const BetaAssistantUpdateParamsModelGPT3_5Turbo1106 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-1106"
    - const BetaAssistantUpdateParamsModelGPT3_5Turbo0125 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-0125"
    - const BetaAssistantUpdateParamsModelGPT3_5Turbo16k0613 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-16k-0613"
- Name param.Field[string]
  
  The name of the assistant. The maximum length is 256 characters.
- ReasoningEffort param.Field[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.
- ResponseFormat param.Field[AssistantResponseFormatOptionUnion]
  
  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.
- Temperature param.Field[float64]
  
  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.
- ToolResources param.Field[BetaAssistantUpdateParamsToolResources]
  
  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.
  - CodeInterpreter BetaAssistantUpdateParamsToolResourcesCodeInterpreter
    - FileIDs []string
      
      Overrides the list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
  - FileSearch BetaAssistantUpdateParamsToolResourcesFileSearch
    - VectorStoreIDs []string
      
      Overrides the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- Tools param.Field[[]any]
  
  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.
- TopP param.Field[float64]
  
  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

type Assistant struct{…}

Represents an assistant that can call the model and use tools.
- ID string
  
  The identifier, which can be referenced in API endpoints.
- CreatedAt int64
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Description string
  
  The description of the assistant. The maximum length is 512 characters.
- Instructions string
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Metadata Metadata
  
  Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
  
  Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
- Model string
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- Name string
  
  The name of the assistant. The maximum length is 256 characters.
- Object Assistant
  
  The object type, which is always assistant.
  - const AssistantAssistant Assistant = "assistant"
- Tools []any
  
  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.
- ResponseFormat AssistantResponseFormatOptionUnion
  
  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.
  - type Auto string
    
    auto is the default value
    - const AutoAuto Auto = "auto"
  - type ResponseFormatText struct{…}
    
    Default response format. Used to generate text responses.
    - Type Text
      
      The type of response format being defined. Always text.
      - const TextText Text = "text"
  - type ResponseFormatJSONObject struct{…}
    
    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 JSONObject
      
      The type of response format being defined. Always json_object.
      - const JSONObjectJSONObject JSONObject = "json_object"
  - type ResponseFormatJSONSchema struct{…}
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JSONSchema ResponseFormatJSONSchemaJSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - Name string
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - Description string
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Schema map[string, any]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Strict 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 JSONSchema
      
      The type of response format being defined. Always json_schema.
      - const JSONSchemaJSONSchema JSONSchema = "json_schema"
- Temperature float64
  
  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.
- ToolResources AssistantToolResources
  
  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.
  - CodeInterpreter AssistantToolResourcesCodeInterpreter
    - FileIDs []string
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - FileSearch AssistantToolResourcesFileSearch
    - VectorStoreIDs []string
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- TopP float64
  
  An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
  
  We generally recommend altering this or temperature but not both.

Example

package main

import (
  "context"
  "fmt"

  "github.com/openai/openai-go"
  "github.com/openai/openai-go/option"
)

func main() {
  client := openai.NewClient(
    option.WithAPIKey("My API Key"),
  )
  assistant, err := client.Beta.Assistants.Update(
    context.TODO(),
    "assistant_id",
    openai.BetaAssistantUpdateParams{

    },
  )
  if err != nil {
    panic(err.Error())
  }
  fmt.Printf("%+v\n", assistant.ID)
}

Response

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

Delete assistant

client.Beta.Assistants.Delete(ctx, assistantID) (*AssistantDeleted, error)

delete /assistants/{assistant_id}

Delete assistant

Parameters

assistantID string

Returns

type AssistantDeleted struct{…}
- ID string
- Deleted bool
- Object AssistantDeleted
  - const AssistantDeletedAssistantDeleted AssistantDeleted = "assistant.deleted"

Example

package main

import (
  "context"
  "fmt"

  "github.com/openai/openai-go"
  "github.com/openai/openai-go/option"
)

func main() {
  client := openai.NewClient(
    option.WithAPIKey("My API Key"),
  )
  assistantDeleted, err := client.Beta.Assistants.Delete(context.TODO(), "assistant_id")
  if err != nil {
    panic(err.Error())
  }
  fmt.Printf("%+v\n", assistantDeleted.ID)
}

Response

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

Domain Types

Assistant

type Assistant struct{…}

Represents an assistant that can call the model and use tools.
- ID string
  
  The identifier, which can be referenced in API endpoints.
- CreatedAt int64
  
  The Unix timestamp (in seconds) for when the assistant was created.
- Description string
  
  The description of the assistant. The maximum length is 512 characters.
- Instructions string
  
  The system instructions that the assistant uses. The maximum length is 256,000 characters.
- Metadata Metadata
  
  Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
  
  Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
- Model string
  
  ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
- Name string
  
  The name of the assistant. The maximum length is 256 characters.
- Object Assistant
  
  The object type, which is always assistant.
  - const AssistantAssistant Assistant = "assistant"
- Tools []any
  
  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.
- ResponseFormat AssistantResponseFormatOptionUnion
  
  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.
  - type Auto string
    
    auto is the default value
    - const AutoAuto Auto = "auto"
  - type ResponseFormatText struct{…}
    
    Default response format. Used to generate text responses.
    - Type Text
      
      The type of response format being defined. Always text.
      - const TextText Text = "text"
  - type ResponseFormatJSONObject struct{…}
    
    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 JSONObject
      
      The type of response format being defined. Always json_object.
      - const JSONObjectJSONObject JSONObject = "json_object"
  - type ResponseFormatJSONSchema struct{…}
    
    JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
    - JSONSchema ResponseFormatJSONSchemaJSONSchema
      
      Structured Outputs configuration options, including a JSON Schema.
      - Name string
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
      - Description string
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
      - Schema map[string, any]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
      - Strict 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 JSONSchema
      
      The type of response format being defined. Always json_schema.
      - const JSONSchemaJSONSchema JSONSchema = "json_schema"
- Temperature float64
  
  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.
- ToolResources AssistantToolResources
  
  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.
  - CodeInterpreter AssistantToolResourcesCodeInterpreter
    - FileIDs []string
      
      A list of file IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.
  - FileSearch AssistantToolResourcesFileSearch
    - VectorStoreIDs []string
      
      The ID of the vector store attached to this assistant. There can be a maximum of 1 vector store attached to the assistant.
- TopP float64
  
  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

type AssistantDeleted struct{…}
- ID string
- Deleted bool
- Object AssistantDeleted
  - const AssistantDeletedAssistantDeleted AssistantDeleted = "assistant.deleted"

Assistant Stream Event

type AssistantStreamEventUnion interface{…}

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.
- type AssistantStreamEventThreadCreated struct{…}
  
  Occurs when a new thread is created.
  - Data Thread
    
    Represents a thread that contains messages.
    - ID string
      
      The identifier, which can be referenced in API endpoints.
    - CreatedAt int64
      
      The Unix timestamp (in seconds) for when the thread was created.
    - Metadata Metadata
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
    - Object Thread
      
      The object type, which is always thread.
      - const ThreadThread Thread = "thread"
    - ToolResources ThreadToolResources
      
      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.
      - CodeInterpreter ThreadToolResourcesCodeInterpreter
        
        FileIDs []string
        
        A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
      - FileSearch ThreadToolResourcesFileSearch
        
        VectorStoreIDs []string
        
        The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.
  - Event ThreadCreated
    - const ThreadCreatedThreadCreated ThreadCreated = "thread.created"
  - Enabled bool
    
    Whether to enable input audio transcription.
- type AssistantStreamEventThreadRunCreated struct{…}
  
  Occurs when a new run is created.
  - Data Run
    
    Represents an execution run on a thread.
    - ID string
      
      The identifier, which can be referenced in API endpoints.
    - AssistantID string
      
      The ID of the assistant used for execution of this run.
    - CancelledAt int64
      
      The Unix timestamp (in seconds) for when the run was cancelled.
    - CompletedAt int64
      
      The Unix timestamp (in seconds) for when the run was completed.
    - CreatedAt int64
      
      The Unix timestamp (in seconds) for when the run was created.
    - ExpiresAt int64
      
      The Unix timestamp (in seconds) for when the run will expire.
    - FailedAt int64
      
      The Unix timestamp (in seconds) for when the run failed.
    - IncompleteDetails RunIncompleteDetails
      
      Details on why the run is incomplete. Will be null if the run is not incomplete.
      - Reason string
        
        The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.
        
        const RunIncompleteDetailsReasonMaxCompletionTokens RunIncompleteDetailsReason = "max_completion_tokens"
        
        const RunIncompleteDetailsReasonMaxPromptTokens RunIncompleteDetailsReason = "max_prompt_tokens"
    - Instructions string
      
      The instructions that the assistant used for this run.
    - LastError RunLastError
      
      The last error associated with this run. Will be null if there are no errors.
      - Code string
        
        One of server_error, rate_limit_exceeded, or invalid_prompt.
        
        const RunLastErrorCodeServerError RunLastErrorCode = "server_error"
        
        const RunLastErrorCodeRateLimitExceeded RunLastErrorCode = "rate_limit_exceeded"
        
        const RunLastErrorCodeInvalidPrompt RunLastErrorCode = "invalid_prompt"
      - Message string
        
        A human-readable description of the error.
    - MaxCompletionTokens int64
      
      The maximum number of completion tokens specified to have been used over the course of the run.
    - MaxPromptTokens int64
      
      The maximum number of prompt tokens specified to have been used over the course of the run.
    - Metadata Metadata
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
    - Model string
      
      The model that the assistant used for this run.
    - Object ThreadRun
      
      The object type, which is always thread.run.
      - const ThreadRunThreadRun ThreadRun = "thread.run"
    - ParallelToolCalls bool
      
      Whether to enable parallel function calling during tool use.
    - RequiredAction RunRequiredAction
      
      Details on the action required to continue the run. Will be null if no action is required.
      - SubmitToolOutputs RunRequiredActionSubmitToolOutputs
        
        Details on the tool outputs needed for this run to continue.
        
        ToolCalls []RequiredActionFunctionToolCall
        
        A list of the relevant tool calls.
        
        ID string
        
        The ID of the tool call. This ID must be referenced when you submit the tool outputs in using the Submit tool outputs to run endpoint.
        
        Function RequiredActionFunctionToolCallFunction
        
        The function definition.
        
        Arguments string
        
        The arguments that the model expects you to pass to the function.
        
        Name string
        
        The name of the function.
        
        Type Function
        
        The type of tool call the output is required for. For now, this is always function.
        
        const FunctionFunction Function = "function"
      - Type SubmitToolOutputs
        
        For now, this is always submit_tool_outputs.
        
        const SubmitToolOutputsSubmitToolOutputs SubmitToolOutputs = "submit_tool_outputs"
    - ResponseFormat AssistantResponseFormatOptionUnion
      
      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.
      - type Auto string
        
        auto is the default value
        
        const AutoAuto Auto = "auto"
      - type ResponseFormatText struct{…}
        
        Default response format. Used to generate text responses.
        
        Type Text
        
        The type of response format being defined. Always text.
        
        const TextText Text = "text"
      - type ResponseFormatJSONObject struct{…}
        
        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 JSONObject
        
        The type of response format being defined. Always json_object.
        
        const JSONObjectJSONObject JSONObject = "json_object"
      - type ResponseFormatJSONSchema struct{…}
        
        JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
        
        JSONSchema ResponseFormatJSONSchemaJSONSchema
        
        Structured Outputs configuration options, including a JSON Schema.
        
        Name string
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
        
        Description string
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
        
        Schema map[string, any]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
        
        Strict 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 JSONSchema
        
        The type of response format being defined. Always json_schema.
        
        const JSONSchemaJSONSchema JSONSchema = "json_schema"
    - StartedAt int64
      
      The Unix timestamp (in seconds) for when the run was started.
    - Status any
    - ThreadID string
      
      The ID of the thread that was executed on as a part of this run.
    - ToolChoice AssistantToolChoiceOptionUnion
      
      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.
      - type AssistantToolChoiceOptionAuto string
        
        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.
        
        const AssistantToolChoiceOptionAutoNone AssistantToolChoiceOptionAuto = "none"
        
        const AssistantToolChoiceOptionAutoAuto AssistantToolChoiceOptionAuto = "auto"
        
        const AssistantToolChoiceOptionAutoRequired AssistantToolChoiceOptionAuto = "required"
      - type AssistantToolChoice struct{…}
        
        Specifies a tool the model should use. Use to force the model to call a specific tool.
        
        Type AssistantToolChoiceType
        
        The type of the tool. If type is function, the function name must be set
        
        const AssistantToolChoiceTypeFunction AssistantToolChoiceType = "function"
        
        const AssistantToolChoiceTypeCodeInterpreter AssistantToolChoiceType = "code_interpreter"
        
        const AssistantToolChoiceTypeFileSearch AssistantToolChoiceType = "file_search"
        
        Function AssistantToolChoiceFunction
        
        Name string
        
        The name of the function to call.
    - Tools []any
      
      The list of tools that the assistant used for this run.
    - TruncationStrategy RunTruncationStrategy
      
      Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.
      - Type string
        
        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.
        
        const RunTruncationStrategyTypeAuto RunTruncationStrategyType = "auto"
        
        const RunTruncationStrategyTypeLastMessages RunTruncationStrategyType = "last_messages"
      - LastMessages int64
        
        The number of most recent messages from the thread when constructing the context for the run.
    - Usage RunUsage
      
      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.).
      - CompletionTokens int64
        
        Number of completion tokens used over the course of the run.
      - PromptTokens int64
        
        Number of prompt tokens used over the course of the run.
      - TotalTokens int64
        
        Total number of tokens used (prompt + completion).
    - Temperature float64
      
      The sampling temperature used for this run. If not set, defaults to 1.
    - TopP float64
      
      The nucleus sampling value used for this run. If not set, defaults to 1.
  - Event ThreadRunCreated
    - const ThreadRunCreatedThreadRunCreated ThreadRunCreated = "thread.run.created"
- type AssistantStreamEventThreadRunQueued struct{…}
  
  Occurs when a run moves to a queued status.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunQueued
    - const ThreadRunQueuedThreadRunQueued ThreadRunQueued = "thread.run.queued"
- type AssistantStreamEventThreadRunInProgress struct{…}
  
  Occurs when a run moves to an in_progress status.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunInProgress
    - const ThreadRunInProgressThreadRunInProgress ThreadRunInProgress = "thread.run.in_progress"
- type AssistantStreamEventThreadRunRequiresAction struct{…}
  
  Occurs when a run moves to a requires_action status.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunRequiresAction
    - const ThreadRunRequiresActionThreadRunRequiresAction ThreadRunRequiresAction = "thread.run.requires_action"
- type AssistantStreamEventThreadRunCompleted struct{…}
  
  Occurs when a run is completed.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunCompleted
    - const ThreadRunCompletedThreadRunCompleted ThreadRunCompleted = "thread.run.completed"
- type AssistantStreamEventThreadRunIncomplete struct{…}
  
  Occurs when a run ends with status incomplete.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunIncomplete
    - const ThreadRunIncompleteThreadRunIncomplete ThreadRunIncomplete = "thread.run.incomplete"
- type AssistantStreamEventThreadRunFailed struct{…}
  
  Occurs when a run fails.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunFailed
    - const ThreadRunFailedThreadRunFailed ThreadRunFailed = "thread.run.failed"
- type AssistantStreamEventThreadRunCancelling struct{…}
  
  Occurs when a run moves to a cancelling status.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunCancelling
    - const ThreadRunCancellingThreadRunCancelling ThreadRunCancelling = "thread.run.cancelling"
- type AssistantStreamEventThreadRunCancelled struct{…}
  
  Occurs when a run is cancelled.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunCancelled
    - const ThreadRunCancelledThreadRunCancelled ThreadRunCancelled = "thread.run.cancelled"
- type AssistantStreamEventThreadRunExpired struct{…}
  
  Occurs when a run expires.
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunExpired
    - const ThreadRunExpiredThreadRunExpired ThreadRunExpired = "thread.run.expired"
- type AssistantStreamEventThreadRunStepCreated struct{…}
  
  Occurs when a run step is created.
  - Data RunStep
    
    Represents a step in execution of a run.
    - ID string
      
      The identifier of the run step, which can be referenced in API endpoints.
    - AssistantID string
      
      The ID of the assistant associated with the run step.
    - CancelledAt int64
      
      The Unix timestamp (in seconds) for when the run step was cancelled.
    - CompletedAt int64
      
      The Unix timestamp (in seconds) for when the run step completed.
    - CreatedAt int64
      
      The Unix timestamp (in seconds) for when the run step was created.
    - ExpiredAt int64
      
      The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.
    - FailedAt int64
      
      The Unix timestamp (in seconds) for when the run step failed.
    - LastError RunStepLastError
      
      The last error associated with this run step. Will be null if there are no errors.
      - Code string
        
        One of server_error or rate_limit_exceeded.
        
        const RunStepLastErrorCodeServerError RunStepLastErrorCode = "server_error"
        
        const RunStepLastErrorCodeRateLimitExceeded RunStepLastErrorCode = "rate_limit_exceeded"
      - Message string
        
        A human-readable description of the error.
    - Metadata Metadata
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
    - Object ThreadRunStep
      
      The object type, which is always thread.run.step.
      - const ThreadRunStepThreadRunStep ThreadRunStep = "thread.run.step"
    - RunID string
      
      The ID of the run that this run step is a part of.
    - Status RunStepStatus
      
      The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.
      - const RunStepStatusInProgress RunStepStatus = "in_progress"
      - const RunStepStatusCancelled RunStepStatus = "cancelled"
      - const RunStepStatusFailed RunStepStatus = "failed"
      - const RunStepStatusCompleted RunStepStatus = "completed"
      - const RunStepStatusExpired RunStepStatus = "expired"
    - StepDetails RunStepStepDetailsUnion
      
      The details of the run step.
      - type MessageCreationStepDetails struct{…}
        
        Details of the message creation by the run step.
        
        MessageCreation MessageCreationStepDetailsMessageCreation
        
        MessageID string
        
        The ID of the message that was created by this run step.
        
        Type MessageCreation
        
        Always message_creation.
        
        const MessageCreationMessageCreation MessageCreation = "message_creation"
      - type ToolCallsStepDetails struct{…}
        
        Details of the tool call.
        
        ToolCalls []any
        
        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.
        
        Type ToolCalls
        
        Always tool_calls.
        
        const ToolCallsToolCalls ToolCalls = "tool_calls"
    - ThreadID string
      
      The ID of the thread that was run.
    - Type RunStepType
      
      The type of run step, which can be either message_creation or tool_calls.
      - const RunStepTypeMessageCreation RunStepType = "message_creation"
      - const RunStepTypeToolCalls RunStepType = "tool_calls"
    - Usage RunStepUsage
      
      Usage statistics related to the run step. This value will be null while the run step's status is in_progress.
      - CompletionTokens int64
        
        Number of completion tokens used over the course of the run step.
      - PromptTokens int64
        
        Number of prompt tokens used over the course of the run step.
      - TotalTokens int64
        
        Total number of tokens used (prompt + completion).
  - Event ThreadRunStepCreated
    - const ThreadRunStepCreatedThreadRunStepCreated ThreadRunStepCreated = "thread.run.step.created"
- type AssistantStreamEventThreadRunStepInProgress struct{…}
  
  Occurs when a run step moves to an in_progress state.
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepInProgress
    - const ThreadRunStepInProgressThreadRunStepInProgress ThreadRunStepInProgress = "thread.run.step.in_progress"
- type AssistantStreamEventThreadRunStepDelta struct{…}
  
  Occurs when parts of a run step are being streamed.
  - Data RunStepDeltaEvent
    
    Represents a run step delta i.e. any changed fields on a run step during streaming.
    - ID string
      
      The identifier of the run step, which can be referenced in API endpoints.
    - Delta any
    - Object ThreadRunStepDelta
      
      The object type, which is always thread.run.step.delta.
      - const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"
  - Event ThreadRunStepDelta
    - const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"
- type AssistantStreamEventThreadRunStepCompleted struct{…}
  
  Occurs when a run step is completed.
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepCompleted
    - const ThreadRunStepCompletedThreadRunStepCompleted ThreadRunStepCompleted = "thread.run.step.completed"
- type AssistantStreamEventThreadRunStepFailed struct{…}
  
  Occurs when a run step fails.
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepFailed
    - const ThreadRunStepFailedThreadRunStepFailed ThreadRunStepFailed = "thread.run.step.failed"
- type AssistantStreamEventThreadRunStepCancelled struct{…}
  
  Occurs when a run step is cancelled.
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepCancelled
    - const ThreadRunStepCancelledThreadRunStepCancelled ThreadRunStepCancelled = "thread.run.step.cancelled"
- type AssistantStreamEventThreadRunStepExpired struct{…}
  
  Occurs when a run step expires.
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepExpired
    - const ThreadRunStepExpiredThreadRunStepExpired ThreadRunStepExpired = "thread.run.step.expired"
- type AssistantStreamEventThreadMessageCreated struct{…}
  
  Occurs when a message is created.
  - Data Message
    
    Represents a message within a thread.
    - ID string
      
      The identifier, which can be referenced in API endpoints.
    - AssistantID string
      
      If applicable, the ID of the assistant that authored this message.
    - Attachments []MessageAttachment
      
      A list of files attached to the message, and the tools they were added to.
      - FileID string
        
        The ID of the file to attach to the message.
      - Tools []MessageAttachmentToolUnion
        
        The tools to add this file to.
        
        type CodeInterpreterTool struct{…}
        
        Type CodeInterpreter
        
        The type of tool being defined: code_interpreter
        
        const CodeInterpreterCodeInterpreter CodeInterpreter = "code_interpreter"
        
        type MessageAttachmentToolAssistantToolsFileSearchTypeOnly struct{…}
        
        Type FileSearch
        
        The type of tool being defined: file_search
        
        const FileSearchFileSearch FileSearch = "file_search"
    - CompletedAt int64
      
      The Unix timestamp (in seconds) for when the message was completed.
    - Content []any
      
      The content of the message in array of text and/or images.
    - CreatedAt int64
      
      The Unix timestamp (in seconds) for when the message was created.
    - IncompleteAt int64
      
      The Unix timestamp (in seconds) for when the message was marked as incomplete.
    - IncompleteDetails MessageIncompleteDetails
      
      On an incomplete message, details about why the message is incomplete.
      - Reason string
        
        The reason the message is incomplete.
        
        const MessageIncompleteDetailsReasonContentFilter MessageIncompleteDetailsReason = "content_filter"
        
        const MessageIncompleteDetailsReasonMaxTokens MessageIncompleteDetailsReason = "max_tokens"
        
        const MessageIncompleteDetailsReasonRunCancelled MessageIncompleteDetailsReason = "run_cancelled"
        
        const MessageIncompleteDetailsReasonRunExpired MessageIncompleteDetailsReason = "run_expired"
        
        const MessageIncompleteDetailsReasonRunFailed MessageIncompleteDetailsReason = "run_failed"
    - Metadata Metadata
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
    - Object ThreadMessage
      
      The object type, which is always thread.message.
      - const ThreadMessageThreadMessage ThreadMessage = "thread.message"
    - Role MessageRole
      
      The entity that produced the message. One of user or assistant.
      - const MessageRoleUser MessageRole = "user"
      - const MessageRoleAssistant MessageRole = "assistant"
    - RunID string
      
      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 MessageStatus
      
      The status of the message, which can be either in_progress, incomplete, or completed.
      - const MessageStatusInProgress MessageStatus = "in_progress"
      - const MessageStatusIncomplete MessageStatus = "incomplete"
      - const MessageStatusCompleted MessageStatus = "completed"
    - ThreadID string
      
      The thread ID that this message belongs to.
  - Event ThreadMessageCreated
    - const ThreadMessageCreatedThreadMessageCreated ThreadMessageCreated = "thread.message.created"
- type AssistantStreamEventThreadMessageInProgress struct{…}
  
  Occurs when a message moves to an in_progress state.
  - Data Message
    
    Represents a message within a thread.
  - Event ThreadMessageInProgress
    - const ThreadMessageInProgressThreadMessageInProgress ThreadMessageInProgress = "thread.message.in_progress"
- type AssistantStreamEventThreadMessageDelta struct{…}
  
  Occurs when parts of a Message are being streamed.
  - Data MessageDeltaEvent
    
    Represents a message delta i.e. any changed fields on a message during streaming.
    - ID string
      
      The identifier of the message, which can be referenced in API endpoints.
    - Delta MessageDelta
      
      The delta containing the fields that have changed on the Message.
      - Content []any
        
        The content of the message in array of text and/or images.
      - Role MessageDeltaRole
        
        The entity that produced the message. One of user or assistant.
        
        const MessageDeltaRoleUser MessageDeltaRole = "user"
        
        const MessageDeltaRoleAssistant MessageDeltaRole = "assistant"
    - Object ThreadMessageDelta
      
      The object type, which is always thread.message.delta.
      - const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"
  - Event ThreadMessageDelta
    - const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"
- type AssistantStreamEventThreadMessageCompleted struct{…}
  
  Occurs when a message is completed.
  - Data Message
    
    Represents a message within a thread.
  - Event ThreadMessageCompleted
    - const ThreadMessageCompletedThreadMessageCompleted ThreadMessageCompleted = "thread.message.completed"
- type AssistantStreamEventThreadMessageIncomplete struct{…}
  
  Occurs when a message ends before it is completed.
  - Data Message
    
    Represents a message within a thread.
  - Event ThreadMessageIncomplete
    - const ThreadMessageIncompleteThreadMessageIncomplete ThreadMessageIncomplete = "thread.message.incomplete"
- type AssistantStreamEventErrorEvent struct{…}
  
  Occurs when an error occurs. This can happen due to an internal server error or a timeout.
  - Data ErrorObject
    - Code string
    - Message string
    - Param string
    - Type string
  - Event Error
    - const ErrorError Error = "error"

Assistant Tool

type AssistantTool interface{…}

Code Interpreter Tool

type CodeInterpreterTool struct{…}
- Type CodeInterpreter
  
  The type of tool being defined: code_interpreter
  - const CodeInterpreterCodeInterpreter CodeInterpreter = "code_interpreter"

File Search Tool

type FileSearchTool struct{…}
- Type FileSearch
  
  The type of tool being defined: file_search
  - const FileSearchFileSearch FileSearch = "file_search"
- FileSearch FileSearchToolFileSearch
  
  Overrides for the file search tool.
  - MaxNumResults int64
    
    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.
  - RankingOptions FileSearchToolFileSearchRankingOptions
    
    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.
    - ScoreThreshold float64
      
      The score threshold for the file search. All values must be a floating point number between 0 and 1.
    - Ranker string
      
      The ranker to use for the file search. If not specified will use the auto ranker.
      - const FileSearchToolFileSearchRankingOptionsRankerAuto FileSearchToolFileSearchRankingOptionsRanker = "auto"
      - const FileSearchToolFileSearchRankingOptionsRankerDefault2024_08_21 FileSearchToolFileSearchRankingOptionsRanker = "default_2024_08_21"

Function Tool

type FunctionTool struct{…}
- Function FunctionDefinition
  - Name string
    
    The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
  - Description string
    
    A description of what the function does, used by the model to choose when and how to call the function.
  - Parameters 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 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 Function
  
  The type of tool being defined: function
  - const FunctionFunction Function = "function"

Message Stream Event

type MessageStreamEventUnion interface{…}

Occurs when a message is created.
- MessageStreamEventThreadMessageCreated
  - Data Message
    
    Represents a message within a thread.
    - ID string
      
      The identifier, which can be referenced in API endpoints.
    - AssistantID string
      
      If applicable, the ID of the assistant that authored this message.
    - Attachments []MessageAttachment
      
      A list of files attached to the message, and the tools they were added to.
      - FileID string
        
        The ID of the file to attach to the message.
      - Tools []MessageAttachmentToolUnion
        
        The tools to add this file to.
        
        type CodeInterpreterTool struct{…}
        
        Type CodeInterpreter
        
        The type of tool being defined: code_interpreter
        
        const CodeInterpreterCodeInterpreter CodeInterpreter = "code_interpreter"
        
        type MessageAttachmentToolAssistantToolsFileSearchTypeOnly struct{…}
        
        Type FileSearch
        
        The type of tool being defined: file_search
        
        const FileSearchFileSearch FileSearch = "file_search"
    - CompletedAt int64
      
      The Unix timestamp (in seconds) for when the message was completed.
    - Content []any
      
      The content of the message in array of text and/or images.
    - CreatedAt int64
      
      The Unix timestamp (in seconds) for when the message was created.
    - IncompleteAt int64
      
      The Unix timestamp (in seconds) for when the message was marked as incomplete.
    - IncompleteDetails MessageIncompleteDetails
      
      On an incomplete message, details about why the message is incomplete.
      - Reason string
        
        The reason the message is incomplete.
        
        const MessageIncompleteDetailsReasonContentFilter MessageIncompleteDetailsReason = "content_filter"
        
        const MessageIncompleteDetailsReasonMaxTokens MessageIncompleteDetailsReason = "max_tokens"
        
        const MessageIncompleteDetailsReasonRunCancelled MessageIncompleteDetailsReason = "run_cancelled"
        
        const MessageIncompleteDetailsReasonRunExpired MessageIncompleteDetailsReason = "run_expired"
        
        const MessageIncompleteDetailsReasonRunFailed MessageIncompleteDetailsReason = "run_failed"
    - Metadata Metadata
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
    - Object ThreadMessage
      
      The object type, which is always thread.message.
      - const ThreadMessageThreadMessage ThreadMessage = "thread.message"
    - Role MessageRole
      
      The entity that produced the message. One of user or assistant.
      - const MessageRoleUser MessageRole = "user"
      - const MessageRoleAssistant MessageRole = "assistant"
    - RunID string
      
      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 MessageStatus
      
      The status of the message, which can be either in_progress, incomplete, or completed.
      - const MessageStatusInProgress MessageStatus = "in_progress"
      - const MessageStatusIncomplete MessageStatus = "incomplete"
      - const MessageStatusCompleted MessageStatus = "completed"
    - ThreadID string
      
      The thread ID that this message belongs to.
  - Event ThreadMessageCreated
    - const ThreadMessageCreatedThreadMessageCreated ThreadMessageCreated = "thread.message.created"
- MessageStreamEventThreadMessageInProgress
  - Data Message
    
    Represents a message within a thread.
  - Event ThreadMessageInProgress
    - const ThreadMessageInProgressThreadMessageInProgress ThreadMessageInProgress = "thread.message.in_progress"
- MessageStreamEventThreadMessageDelta
  - Data MessageDeltaEvent
    
    Represents a message delta i.e. any changed fields on a message during streaming.
    - ID string
      
      The identifier of the message, which can be referenced in API endpoints.
    - Delta MessageDelta
      
      The delta containing the fields that have changed on the Message.
      - Content []any
        
        The content of the message in array of text and/or images.
      - Role MessageDeltaRole
        
        The entity that produced the message. One of user or assistant.
        
        const MessageDeltaRoleUser MessageDeltaRole = "user"
        
        const MessageDeltaRoleAssistant MessageDeltaRole = "assistant"
    - Object ThreadMessageDelta
      
      The object type, which is always thread.message.delta.
      - const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"
  - Event ThreadMessageDelta
    - const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"
- MessageStreamEventThreadMessageCompleted
  - Data Message
    
    Represents a message within a thread.
  - Event ThreadMessageCompleted
    - const ThreadMessageCompletedThreadMessageCompleted ThreadMessageCompleted = "thread.message.completed"
- MessageStreamEventThreadMessageIncomplete
  - Data Message
    
    Represents a message within a thread.
  - Event ThreadMessageIncomplete
    - const ThreadMessageIncompleteThreadMessageIncomplete ThreadMessageIncomplete = "thread.message.incomplete"

Run Step Stream Event

type RunStepStreamEventUnion interface{…}

Occurs when a run step is created.
- RunStepStreamEventThreadRunStepCreated
  - Data RunStep
    
    Represents a step in execution of a run.
    - ID string
      
      The identifier of the run step, which can be referenced in API endpoints.
    - AssistantID string
      
      The ID of the assistant associated with the run step.
    - CancelledAt int64
      
      The Unix timestamp (in seconds) for when the run step was cancelled.
    - CompletedAt int64
      
      The Unix timestamp (in seconds) for when the run step completed.
    - CreatedAt int64
      
      The Unix timestamp (in seconds) for when the run step was created.
    - ExpiredAt int64
      
      The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.
    - FailedAt int64
      
      The Unix timestamp (in seconds) for when the run step failed.
    - LastError RunStepLastError
      
      The last error associated with this run step. Will be null if there are no errors.
      - Code string
        
        One of server_error or rate_limit_exceeded.
        
        const RunStepLastErrorCodeServerError RunStepLastErrorCode = "server_error"
        
        const RunStepLastErrorCodeRateLimitExceeded RunStepLastErrorCode = "rate_limit_exceeded"
      - Message string
        
        A human-readable description of the error.
    - Metadata Metadata
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
    - Object ThreadRunStep
      
      The object type, which is always thread.run.step.
      - const ThreadRunStepThreadRunStep ThreadRunStep = "thread.run.step"
    - RunID string
      
      The ID of the run that this run step is a part of.
    - Status RunStepStatus
      
      The status of the run step, which can be either in_progress, cancelled, failed, completed, or expired.
      - const RunStepStatusInProgress RunStepStatus = "in_progress"
      - const RunStepStatusCancelled RunStepStatus = "cancelled"
      - const RunStepStatusFailed RunStepStatus = "failed"
      - const RunStepStatusCompleted RunStepStatus = "completed"
      - const RunStepStatusExpired RunStepStatus = "expired"
    - StepDetails RunStepStepDetailsUnion
      
      The details of the run step.
      - type MessageCreationStepDetails struct{…}
        
        Details of the message creation by the run step.
        
        MessageCreation MessageCreationStepDetailsMessageCreation
        
        MessageID string
        
        The ID of the message that was created by this run step.
        
        Type MessageCreation
        
        Always message_creation.
        
        const MessageCreationMessageCreation MessageCreation = "message_creation"
      - type ToolCallsStepDetails struct{…}
        
        Details of the tool call.
        
        ToolCalls []any
        
        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.
        
        Type ToolCalls
        
        Always tool_calls.
        
        const ToolCallsToolCalls ToolCalls = "tool_calls"
    - ThreadID string
      
      The ID of the thread that was run.
    - Type RunStepType
      
      The type of run step, which can be either message_creation or tool_calls.
      - const RunStepTypeMessageCreation RunStepType = "message_creation"
      - const RunStepTypeToolCalls RunStepType = "tool_calls"
    - Usage RunStepUsage
      
      Usage statistics related to the run step. This value will be null while the run step's status is in_progress.
      - CompletionTokens int64
        
        Number of completion tokens used over the course of the run step.
      - PromptTokens int64
        
        Number of prompt tokens used over the course of the run step.
      - TotalTokens int64
        
        Total number of tokens used (prompt + completion).
  - Event ThreadRunStepCreated
    - const ThreadRunStepCreatedThreadRunStepCreated ThreadRunStepCreated = "thread.run.step.created"
- RunStepStreamEventThreadRunStepInProgress
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepInProgress
    - const ThreadRunStepInProgressThreadRunStepInProgress ThreadRunStepInProgress = "thread.run.step.in_progress"
- RunStepStreamEventThreadRunStepDelta
  - Data RunStepDeltaEvent
    
    Represents a run step delta i.e. any changed fields on a run step during streaming.
    - ID string
      
      The identifier of the run step, which can be referenced in API endpoints.
    - Delta any
    - Object ThreadRunStepDelta
      
      The object type, which is always thread.run.step.delta.
      - const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"
  - Event ThreadRunStepDelta
    - const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"
- RunStepStreamEventThreadRunStepCompleted
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepCompleted
    - const ThreadRunStepCompletedThreadRunStepCompleted ThreadRunStepCompleted = "thread.run.step.completed"
- RunStepStreamEventThreadRunStepFailed
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepFailed
    - const ThreadRunStepFailedThreadRunStepFailed ThreadRunStepFailed = "thread.run.step.failed"
- RunStepStreamEventThreadRunStepCancelled
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepCancelled
    - const ThreadRunStepCancelledThreadRunStepCancelled ThreadRunStepCancelled = "thread.run.step.cancelled"
- RunStepStreamEventThreadRunStepExpired
  - Data RunStep
    
    Represents a step in execution of a run.
  - Event ThreadRunStepExpired
    - const ThreadRunStepExpiredThreadRunStepExpired ThreadRunStepExpired = "thread.run.step.expired"

Run Stream Event

type RunStreamEventUnion interface{…}

Occurs when a new run is created.
- RunStreamEventThreadRunCreated
  - Data Run
    
    Represents an execution run on a thread.
    - ID string
      
      The identifier, which can be referenced in API endpoints.
    - AssistantID string
      
      The ID of the assistant used for execution of this run.
    - CancelledAt int64
      
      The Unix timestamp (in seconds) for when the run was cancelled.
    - CompletedAt int64
      
      The Unix timestamp (in seconds) for when the run was completed.
    - CreatedAt int64
      
      The Unix timestamp (in seconds) for when the run was created.
    - ExpiresAt int64
      
      The Unix timestamp (in seconds) for when the run will expire.
    - FailedAt int64
      
      The Unix timestamp (in seconds) for when the run failed.
    - IncompleteDetails RunIncompleteDetails
      
      Details on why the run is incomplete. Will be null if the run is not incomplete.
      - Reason string
        
        The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.
        
        const RunIncompleteDetailsReasonMaxCompletionTokens RunIncompleteDetailsReason = "max_completion_tokens"
        
        const RunIncompleteDetailsReasonMaxPromptTokens RunIncompleteDetailsReason = "max_prompt_tokens"
    - Instructions string
      
      The instructions that the assistant used for this run.
    - LastError RunLastError
      
      The last error associated with this run. Will be null if there are no errors.
      - Code string
        
        One of server_error, rate_limit_exceeded, or invalid_prompt.
        
        const RunLastErrorCodeServerError RunLastErrorCode = "server_error"
        
        const RunLastErrorCodeRateLimitExceeded RunLastErrorCode = "rate_limit_exceeded"
        
        const RunLastErrorCodeInvalidPrompt RunLastErrorCode = "invalid_prompt"
      - Message string
        
        A human-readable description of the error.
    - MaxCompletionTokens int64
      
      The maximum number of completion tokens specified to have been used over the course of the run.
    - MaxPromptTokens int64
      
      The maximum number of prompt tokens specified to have been used over the course of the run.
    - Metadata Metadata
      
      Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
      
      Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
    - Model string
      
      The model that the assistant used for this run.
    - Object ThreadRun
      
      The object type, which is always thread.run.
      - const ThreadRunThreadRun ThreadRun = "thread.run"
    - ParallelToolCalls bool
      
      Whether to enable parallel function calling during tool use.
    - RequiredAction RunRequiredAction
      
      Details on the action required to continue the run. Will be null if no action is required.
      - SubmitToolOutputs RunRequiredActionSubmitToolOutputs
        
        Details on the tool outputs needed for this run to continue.
        
        ToolCalls []RequiredActionFunctionToolCall
        
        A list of the relevant tool calls.
        
        ID string
        
        The ID of the tool call. This ID must be referenced when you submit the tool outputs in using the Submit tool outputs to run endpoint.
        
        Function RequiredActionFunctionToolCallFunction
        
        The function definition.
        
        Arguments string
        
        The arguments that the model expects you to pass to the function.
        
        Name string
        
        The name of the function.
        
        Type Function
        
        The type of tool call the output is required for. For now, this is always function.
        
        const FunctionFunction Function = "function"
      - Type SubmitToolOutputs
        
        For now, this is always submit_tool_outputs.
        
        const SubmitToolOutputsSubmitToolOutputs SubmitToolOutputs = "submit_tool_outputs"
    - ResponseFormat AssistantResponseFormatOptionUnion
      
      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.
      - type Auto string
        
        auto is the default value
        
        const AutoAuto Auto = "auto"
      - type ResponseFormatText struct{…}
        
        Default response format. Used to generate text responses.
        
        Type Text
        
        The type of response format being defined. Always text.
        
        const TextText Text = "text"
      - type ResponseFormatJSONObject struct{…}
        
        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 JSONObject
        
        The type of response format being defined. Always json_object.
        
        const JSONObjectJSONObject JSONObject = "json_object"
      - type ResponseFormatJSONSchema struct{…}
        
        JSON Schema response format. Used to generate structured JSON responses. Learn more about Structured Outputs.
        
        JSONSchema ResponseFormatJSONSchemaJSONSchema
        
        Structured Outputs configuration options, including a JSON Schema.
        
        Name string
        
        The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
        
        Description string
        
        A description of what the response format is for, used by the model to determine how to respond in the format.
        
        Schema map[string, any]
        
        The schema for the response format, described as a JSON Schema object. Learn how to build JSON schemas here.
        
        Strict 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 JSONSchema
        
        The type of response format being defined. Always json_schema.
        
        const JSONSchemaJSONSchema JSONSchema = "json_schema"
    - StartedAt int64
      
      The Unix timestamp (in seconds) for when the run was started.
    - Status any
    - ThreadID string
      
      The ID of the thread that was executed on as a part of this run.
    - ToolChoice AssistantToolChoiceOptionUnion
      
      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.
      - type AssistantToolChoiceOptionAuto string
        
        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.
        
        const AssistantToolChoiceOptionAutoNone AssistantToolChoiceOptionAuto = "none"
        
        const AssistantToolChoiceOptionAutoAuto AssistantToolChoiceOptionAuto = "auto"
        
        const AssistantToolChoiceOptionAutoRequired AssistantToolChoiceOptionAuto = "required"
      - type AssistantToolChoice struct{…}
        
        Specifies a tool the model should use. Use to force the model to call a specific tool.
        
        Type AssistantToolChoiceType
        
        The type of the tool. If type is function, the function name must be set
        
        const AssistantToolChoiceTypeFunction AssistantToolChoiceType = "function"
        
        const AssistantToolChoiceTypeCodeInterpreter AssistantToolChoiceType = "code_interpreter"
        
        const AssistantToolChoiceTypeFileSearch AssistantToolChoiceType = "file_search"
        
        Function AssistantToolChoiceFunction
        
        Name string
        
        The name of the function to call.
    - Tools []any
      
      The list of tools that the assistant used for this run.
    - TruncationStrategy RunTruncationStrategy
      
      Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.
      - Type string
        
        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.
        
        const RunTruncationStrategyTypeAuto RunTruncationStrategyType = "auto"
        
        const RunTruncationStrategyTypeLastMessages RunTruncationStrategyType = "last_messages"
      - LastMessages int64
        
        The number of most recent messages from the thread when constructing the context for the run.
    - Usage RunUsage
      
      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.).
      - CompletionTokens int64
        
        Number of completion tokens used over the course of the run.
      - PromptTokens int64
        
        Number of prompt tokens used over the course of the run.
      - TotalTokens int64
        
        Total number of tokens used (prompt + completion).
    - Temperature float64
      
      The sampling temperature used for this run. If not set, defaults to 1.
    - TopP float64
      
      The nucleus sampling value used for this run. If not set, defaults to 1.
  - Event ThreadRunCreated
    - const ThreadRunCreatedThreadRunCreated ThreadRunCreated = "thread.run.created"
- RunStreamEventThreadRunQueued
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunQueued
    - const ThreadRunQueuedThreadRunQueued ThreadRunQueued = "thread.run.queued"
- RunStreamEventThreadRunInProgress
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunInProgress
    - const ThreadRunInProgressThreadRunInProgress ThreadRunInProgress = "thread.run.in_progress"
- RunStreamEventThreadRunRequiresAction
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunRequiresAction
    - const ThreadRunRequiresActionThreadRunRequiresAction ThreadRunRequiresAction = "thread.run.requires_action"
- RunStreamEventThreadRunCompleted
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunCompleted
    - const ThreadRunCompletedThreadRunCompleted ThreadRunCompleted = "thread.run.completed"
- RunStreamEventThreadRunIncomplete
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunIncomplete
    - const ThreadRunIncompleteThreadRunIncomplete ThreadRunIncomplete = "thread.run.incomplete"
- RunStreamEventThreadRunFailed
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunFailed
    - const ThreadRunFailedThreadRunFailed ThreadRunFailed = "thread.run.failed"
- RunStreamEventThreadRunCancelling
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunCancelling
    - const ThreadRunCancellingThreadRunCancelling ThreadRunCancelling = "thread.run.cancelling"
- RunStreamEventThreadRunCancelled
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunCancelled
    - const ThreadRunCancelledThreadRunCancelled ThreadRunCancelled = "thread.run.cancelled"
- RunStreamEventThreadRunExpired
  - Data Run
    
    Represents an execution run on a thread.
  - Event ThreadRunExpired
    - const ThreadRunExpiredThreadRunExpired ThreadRunExpired = "thread.run.expired"

Thread Stream Event

type ThreadStreamEvent struct{…}

Occurs when a new thread is created.
- Data Thread
  
  Represents a thread that contains messages.
  - ID string
    
    The identifier, which can be referenced in API endpoints.
  - CreatedAt int64
    
    The Unix timestamp (in seconds) for when the thread was created.
  - Metadata Metadata
    
    Set of 16 key-value pairs that can be attached to an object. This can be useful for storing additional information about the object in a structured format, and querying for objects via API or the dashboard.
    
    Keys are strings with a maximum length of 64 characters. Values are strings with a maximum length of 512 characters.
  - Object Thread
    
    The object type, which is always thread.
    - const ThreadThread Thread = "thread"
  - ToolResources ThreadToolResources
    
    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.
    - CodeInterpreter ThreadToolResourcesCodeInterpreter
      - FileIDs []string
        
        A list of file IDs made available to the code_interpreter tool. There can be a maximum of 20 files associated with the tool.
    - FileSearch ThreadToolResourcesFileSearch
      - VectorStoreIDs []string
        
        The vector store attached to this thread. There can be a maximum of 1 vector store attached to the thread.
- Event ThreadCreated
  - const ThreadCreatedThreadCreated ThreadCreated = "thread.created"
- Enabled bool
  
  Whether to enable input audio transcription.

go/resources/beta/subresources/assistants/index.md +3538 −0 created

1# Assistants

3## List assistants

5`client.Beta.Assistants.List(ctx, query) (*CursorPage[Assistant], error)`

7**get** `/assistants`

9List assistants

11### Parameters

13- `query BetaAssistantListParams`

15 - `After param.Field[string]`

17 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.

19 - `Before param.Field[string]`

21 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.

23 - `Limit param.Field[int64]`

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

27 - `Order param.Field[BetaAssistantListParamsOrder]`

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

31 - `const BetaAssistantListParamsOrderAsc BetaAssistantListParamsOrder = "asc"`

33 - `const BetaAssistantListParamsOrderDesc BetaAssistantListParamsOrder = "desc"`

35### Returns

37- `type Assistant struct{…}`

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

41 - `ID string`

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

45 - `CreatedAt int64`

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

49 - `Description string`

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

53 - `Instructions string`

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

57 - `Metadata Metadata`

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

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

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

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

64 with a maximum length of 512 characters.

66 - `Model string`

68 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

70 - `Name string`

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

74 - `Object Assistant`

76 The object type, which is always `assistant`.

78 - `const AssistantAssistant Assistant = "assistant"`

80 - `Tools []any`

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

84 - `ResponseFormat AssistantResponseFormatOptionUnion`

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

88 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).

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

92 **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.

94 - `type Auto string`

96 `auto` is the default value

98 - `const AutoAuto Auto = "auto"`

100 - `type ResponseFormatText struct{…}`

101

102 Default response format. Used to generate text responses.

103

104 - `Type Text`

105

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

107

108 - `const TextText Text = "text"`

109

110 - `type ResponseFormatJSONObject struct{…}`

111

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

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

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

115 to do so.

116

117 - `Type JSONObject`

118

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

120

121 - `const JSONObjectJSONObject JSONObject = "json_object"`

122

123 - `type ResponseFormatJSONSchema struct{…}`

124

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

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

127

128 - `JSONSchema ResponseFormatJSONSchemaJSONSchema`

129

130 Structured Outputs configuration options, including a JSON Schema.

131

132 - `Name string`

133

134 The name of the response format. Must be a-z, A-Z, 0-9, or contain

135 underscores and dashes, with a maximum length of 64.

136

137 - `Description string`

138

139 A description of what the response format is for, used by the model to

140 determine how to respond in the format.

141

142 - `Schema map[string, any]`

143

144 The schema for the response format, described as a JSON Schema object.

145 Learn how to build JSON schemas [here](https://json-schema.org/).

146

147 - `Strict bool`

148

149 Whether to enable strict schema adherence when generating the output.

150 If set to true, the model will always follow the exact schema defined

151 in the `schema` field. Only a subset of JSON Schema is supported when

152 `strict` is `true`. To learn more, read the [Structured Outputs

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

154

155 - `Type JSONSchema`

156

157 The type of response format being defined. Always `json_schema`.

158

159 - `const JSONSchemaJSONSchema JSONSchema = "json_schema"`

160

161 - `Temperature float64`

162

163 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.

164

165 - `ToolResources AssistantToolResources`

166

167 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.

168

169 - `CodeInterpreter AssistantToolResourcesCodeInterpreter`

170

171 - `FileIDs []string`

172

173 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

174

175 - `FileSearch AssistantToolResourcesFileSearch`

176

177 - `VectorStoreIDs []string`

178

179 The ID of 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.

180

181 - `TopP float64`

182

183 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.

184

185 We generally recommend altering this or temperature but not both.

186

187### Example

188

189```go

190package main

191

192import (

193 "context"

194 "fmt"

195

196 "github.com/openai/openai-go"

197 "github.com/openai/openai-go/option"

198)

199

200func main() {

201 client := openai.NewClient(

202 option.WithAPIKey("My API Key"),

203 )

204 page, err := client.Beta.Assistants.List(context.TODO(), openai.BetaAssistantListParams{

205

206 })

207 if err != nil {

208 panic(err.Error())

209 }

210 fmt.Printf("%+v\n", page)

211}

212```

213

214#### Response

215

216```json

217{

218 "data": [

219 {

220 "id": "id",

221 "created_at": 0,

222 "description": "description",

223 "instructions": "instructions",

224 "metadata": {

225 "foo": "string"

226 },

227 "model": "model",

228 "name": "name",

229 "object": "assistant",

230 "tools": [

231 {}

232 ],

233 "response_format": "auto",

234 "temperature": 1,

235 "tool_resources": {

236 "code_interpreter": {

237 "file_ids": [

238 "string"

239 ]

240 },

241 "file_search": {

242 "vector_store_ids": [

243 "string"

244 ]

245 }

246 },

247 "top_p": 1

248 }

249 ],

250 "first_id": "asst_abc123",

251 "has_more": false,

252 "last_id": "asst_abc456",

253 "object": "list"

254}

255```

256

257## Create assistant

258

259`client.Beta.Assistants.New(ctx, body) (*Assistant, error)`

260

261**post** `/assistants`

262

263Create assistant

264

265### Parameters

266

267- `body BetaAssistantNewParams`

268

269 - `Model param.Field[ChatModel]`

270

271 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

272

273 - `string`

274

275 - `type ChatModel string`

276

277 - `const ChatModelGPT5_6Sol ChatModel = "gpt-5.6-sol"`

278

279 - `const ChatModelGPT5_6Terra ChatModel = "gpt-5.6-terra"`

280

281 - `const ChatModelGPT5_6Luna ChatModel = "gpt-5.6-luna"`

282

283 - `const ChatModelGPT5_4 ChatModel = "gpt-5.4"`

284

285 - `const ChatModelGPT5_4Mini ChatModel = "gpt-5.4-mini"`

286

287 - `const ChatModelGPT5_4Nano ChatModel = "gpt-5.4-nano"`

288

289 - `const ChatModelGPT5_4Mini2026_03_17 ChatModel = "gpt-5.4-mini-2026-03-17"`

290

291 - `const ChatModelGPT5_4Nano2026_03_17 ChatModel = "gpt-5.4-nano-2026-03-17"`

292

293 - `const ChatModelGPT5_3ChatLatest ChatModel = "gpt-5.3-chat-latest"`

294

295 - `const ChatModelGPT5_2 ChatModel = "gpt-5.2"`

296

297 - `const ChatModelGPT5_2_2025_12_11 ChatModel = "gpt-5.2-2025-12-11"`

298

299 - `const ChatModelGPT5_2ChatLatest ChatModel = "gpt-5.2-chat-latest"`

300

301 - `const ChatModelGPT5_2Pro ChatModel = "gpt-5.2-pro"`

302

303 - `const ChatModelGPT5_2Pro2025_12_11 ChatModel = "gpt-5.2-pro-2025-12-11"`

304

305 - `const ChatModelGPT5_1 ChatModel = "gpt-5.1"`

306

307 - `const ChatModelGPT5_1_2025_11_13 ChatModel = "gpt-5.1-2025-11-13"`

308

309 - `const ChatModelGPT5_1Codex ChatModel = "gpt-5.1-codex"`

310

311 - `const ChatModelGPT5_1Mini ChatModel = "gpt-5.1-mini"`

312

313 - `const ChatModelGPT5_1ChatLatest ChatModel = "gpt-5.1-chat-latest"`

314

315 - `const ChatModelGPT5 ChatModel = "gpt-5"`

316

317 - `const ChatModelGPT5Mini ChatModel = "gpt-5-mini"`

318

319 - `const ChatModelGPT5Nano ChatModel = "gpt-5-nano"`

320

321 - `const ChatModelGPT5_2025_08_07 ChatModel = "gpt-5-2025-08-07"`

322

323 - `const ChatModelGPT5Mini2025_08_07 ChatModel = "gpt-5-mini-2025-08-07"`

324

325 - `const ChatModelGPT5Nano2025_08_07 ChatModel = "gpt-5-nano-2025-08-07"`

326

327 - `const ChatModelGPT5ChatLatest ChatModel = "gpt-5-chat-latest"`

328

329 - `const ChatModelGPT4_1 ChatModel = "gpt-4.1"`

330

331 - `const ChatModelGPT4_1Mini ChatModel = "gpt-4.1-mini"`

332

333 - `const ChatModelGPT4_1Nano ChatModel = "gpt-4.1-nano"`

334

335 - `const ChatModelGPT4_1_2025_04_14 ChatModel = "gpt-4.1-2025-04-14"`

336

337 - `const ChatModelGPT4_1Mini2025_04_14 ChatModel = "gpt-4.1-mini-2025-04-14"`

338

339 - `const ChatModelGPT4_1Nano2025_04_14 ChatModel = "gpt-4.1-nano-2025-04-14"`

340

341 - `const ChatModelO4Mini ChatModel = "o4-mini"`

342

343 - `const ChatModelO4Mini2025_04_16 ChatModel = "o4-mini-2025-04-16"`

344

345 - `const ChatModelO3 ChatModel = "o3"`

346

347 - `const ChatModelO3_2025_04_16 ChatModel = "o3-2025-04-16"`

348

349 - `const ChatModelO3Mini ChatModel = "o3-mini"`

350

351 - `const ChatModelO3Mini2025_01_31 ChatModel = "o3-mini-2025-01-31"`

352

353 - `const ChatModelO1 ChatModel = "o1"`

354

355 - `const ChatModelO1_2024_12_17 ChatModel = "o1-2024-12-17"`

356

357 - `const ChatModelO1Preview ChatModel = "o1-preview"`

358

359 - `const ChatModelO1Preview2024_09_12 ChatModel = "o1-preview-2024-09-12"`

360

361 - `const ChatModelO1Mini ChatModel = "o1-mini"`

362

363 - `const ChatModelO1Mini2024_09_12 ChatModel = "o1-mini-2024-09-12"`

364

365 - `const ChatModelGPT4o ChatModel = "gpt-4o"`

366

367 - `const ChatModelGPT4o2024_11_20 ChatModel = "gpt-4o-2024-11-20"`

368

369 - `const ChatModelGPT4o2024_08_06 ChatModel = "gpt-4o-2024-08-06"`

370

371 - `const ChatModelGPT4o2024_05_13 ChatModel = "gpt-4o-2024-05-13"`

372

373 - `const ChatModelGPT4oAudioPreview ChatModel = "gpt-4o-audio-preview"`

374

375 - `const ChatModelGPT4oAudioPreview2024_10_01 ChatModel = "gpt-4o-audio-preview-2024-10-01"`

376

377 - `const ChatModelGPT4oAudioPreview2024_12_17 ChatModel = "gpt-4o-audio-preview-2024-12-17"`

378

379 - `const ChatModelGPT4oAudioPreview2025_06_03 ChatModel = "gpt-4o-audio-preview-2025-06-03"`

380

381 - `const ChatModelGPT4oMiniAudioPreview ChatModel = "gpt-4o-mini-audio-preview"`

382

383 - `const ChatModelGPT4oMiniAudioPreview2024_12_17 ChatModel = "gpt-4o-mini-audio-preview-2024-12-17"`

384

385 - `const ChatModelGPT4oSearchPreview ChatModel = "gpt-4o-search-preview"`

386

387 - `const ChatModelGPT4oMiniSearchPreview ChatModel = "gpt-4o-mini-search-preview"`

388

389 - `const ChatModelGPT4oSearchPreview2025_03_11 ChatModel = "gpt-4o-search-preview-2025-03-11"`

390

391 - `const ChatModelGPT4oMiniSearchPreview2025_03_11 ChatModel = "gpt-4o-mini-search-preview-2025-03-11"`

392

393 - `const ChatModelChatgpt4oLatest ChatModel = "chatgpt-4o-latest"`

394

395 - `const ChatModelCodexMiniLatest ChatModel = "codex-mini-latest"`

396

397 - `const ChatModelGPT4oMini ChatModel = "gpt-4o-mini"`

398

399 - `const ChatModelGPT4oMini2024_07_18 ChatModel = "gpt-4o-mini-2024-07-18"`

400

401 - `const ChatModelGPT4Turbo ChatModel = "gpt-4-turbo"`

402

403 - `const ChatModelGPT4Turbo2024_04_09 ChatModel = "gpt-4-turbo-2024-04-09"`

404

405 - `const ChatModelGPT4_0125Preview ChatModel = "gpt-4-0125-preview"`

406

407 - `const ChatModelGPT4TurboPreview ChatModel = "gpt-4-turbo-preview"`

408

409 - `const ChatModelGPT4_1106Preview ChatModel = "gpt-4-1106-preview"`

410

411 - `const ChatModelGPT4VisionPreview ChatModel = "gpt-4-vision-preview"`

412

413 - `const ChatModelGPT4 ChatModel = "gpt-4"`

414

415 - `const ChatModelGPT4_0314 ChatModel = "gpt-4-0314"`

416

417 - `const ChatModelGPT4_0613 ChatModel = "gpt-4-0613"`

418

419 - `const ChatModelGPT4_32k ChatModel = "gpt-4-32k"`

420

421 - `const ChatModelGPT4_32k0314 ChatModel = "gpt-4-32k-0314"`

422

423 - `const ChatModelGPT4_32k0613 ChatModel = "gpt-4-32k-0613"`

424

425 - `const ChatModelGPT3_5Turbo ChatModel = "gpt-3.5-turbo"`

426

427 - `const ChatModelGPT3_5Turbo16k ChatModel = "gpt-3.5-turbo-16k"`

428

429 - `const ChatModelGPT3_5Turbo0301 ChatModel = "gpt-3.5-turbo-0301"`

430

431 - `const ChatModelGPT3_5Turbo0613 ChatModel = "gpt-3.5-turbo-0613"`

432

433 - `const ChatModelGPT3_5Turbo1106 ChatModel = "gpt-3.5-turbo-1106"`

434

435 - `const ChatModelGPT3_5Turbo0125 ChatModel = "gpt-3.5-turbo-0125"`

436

437 - `const ChatModelGPT3_5Turbo16k0613 ChatModel = "gpt-3.5-turbo-16k-0613"`

438

439 - `Description param.Field[string]`

440

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

442

443 - `Instructions param.Field[string]`

444

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

446

447 - `Metadata param.Field[Metadata]`

448

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

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

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

452

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

454 with a maximum length of 512 characters.

455

456 - `Name param.Field[string]`

457

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

459

460 - `ReasoningEffort param.Field[ReasoningEffort]`

461

462 Constrains effort on reasoning for reasoning models. Currently supported

463 values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`.

464 Reducing reasoning effort can result in faster responses and fewer tokens

465 used on reasoning in a response. Not all reasoning models support every

466 value. See the

467 [reasoning guide](https://platform.openai.com/docs/guides/reasoning)

468 for model-specific support.

469

470 - `ResponseFormat param.Field[AssistantResponseFormatOptionUnion]`

471

472 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`.

473

474 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).

475

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

477

478 **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.

479

480 - `Temperature param.Field[float64]`

481

482 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.

483

484 - `ToolResources param.Field[BetaAssistantNewParamsToolResources]`

485

486 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.

487

488 - `CodeInterpreter BetaAssistantNewParamsToolResourcesCodeInterpreter`

489

490 - `FileIDs []string`

491

492 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter` tool. There can be a maximum of 20 files associated with the tool.

493

494 - `FileSearch BetaAssistantNewParamsToolResourcesFileSearch`

495

496 - `VectorStoreIDs []string`

497

498 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.

499

500 - `VectorStores []BetaAssistantNewParamsToolResourcesFileSearchVectorStore`

501

502 A helper to create a [vector store](https://platform.openai.com/docs/api-reference/vector-stores/object) with file_ids and attach it to this assistant. There can be a maximum of 1 vector store attached to the assistant.

503

504 - `ChunkingStrategy BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyUnion`

505

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

507

508 - `type BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyAuto struct{…}`

509

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

511

512 - `Type Auto`

513

514 Always `auto`.

515

516 - `const AutoAuto Auto = "auto"`

517

518 - `type BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyStatic struct{…}`

519

520 - `Static BetaAssistantNewParamsToolResourcesFileSearchVectorStoreChunkingStrategyStaticStatic`

521

522 - `ChunkOverlapTokens int64`

523

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

525

526 Note that the overlap must not exceed half of `max_chunk_size_tokens`.

527

528 - `MaxChunkSizeTokens int64`

529

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

531

532 - `Type Static`

533

534 Always `static`.

535

536 - `const StaticStatic Static = "static"`

537

538 - `FileIDs []string`

539

540 A list of [file](https://platform.openai.com/docs/api-reference/files) 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.

541

542 - `Metadata Metadata`

543

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

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

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

547

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

549 with a maximum length of 512 characters.

550

551 - `Tools param.Field[[]any]`

552

553 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`.

554

555 - `TopP param.Field[float64]`

556

557 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.

558

559 We generally recommend altering this or temperature but not both.

560

561### Returns

562

563- `type Assistant struct{…}`

564

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

566

567 - `ID string`

568

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

570

571 - `CreatedAt int64`

572

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

574

575 - `Description string`

576

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

578

579 - `Instructions string`

580

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

582

583 - `Metadata Metadata`

584

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

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

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

588

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

590 with a maximum length of 512 characters.

591

592 - `Model string`

593

594 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

595

596 - `Name string`

597

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

599

600 - `Object Assistant`

601

602 The object type, which is always `assistant`.

603

604 - `const AssistantAssistant Assistant = "assistant"`

605

606 - `Tools []any`

607

608 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`.

609

610 - `ResponseFormat AssistantResponseFormatOptionUnion`

611

612 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`.

613

614 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).

615

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

617

618 **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.

619

620 - `type Auto string`

621

622 `auto` is the default value

623

624 - `const AutoAuto Auto = "auto"`

625

626 - `type ResponseFormatText struct{…}`

627

628 Default response format. Used to generate text responses.

629

630 - `Type Text`

631

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

633

634 - `const TextText Text = "text"`

635

636 - `type ResponseFormatJSONObject struct{…}`

637

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

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

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

641 to do so.

642

643 - `Type JSONObject`

644

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

646

647 - `const JSONObjectJSONObject JSONObject = "json_object"`

648

649 - `type ResponseFormatJSONSchema struct{…}`

650

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

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

653

654 - `JSONSchema ResponseFormatJSONSchemaJSONSchema`

655

656 Structured Outputs configuration options, including a JSON Schema.

657

658 - `Name string`

659

660 The name of the response format. Must be a-z, A-Z, 0-9, or contain

661 underscores and dashes, with a maximum length of 64.

662

663 - `Description string`

664

665 A description of what the response format is for, used by the model to

666 determine how to respond in the format.

667

668 - `Schema map[string, any]`

669

670 The schema for the response format, described as a JSON Schema object.

671 Learn how to build JSON schemas [here](https://json-schema.org/).

672

673 - `Strict bool`

674

675 Whether to enable strict schema adherence when generating the output.

676 If set to true, the model will always follow the exact schema defined

677 in the `schema` field. Only a subset of JSON Schema is supported when

678 `strict` is `true`. To learn more, read the [Structured Outputs

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

680

681 - `Type JSONSchema`

682

683 The type of response format being defined. Always `json_schema`.

684

685 - `const JSONSchemaJSONSchema JSONSchema = "json_schema"`

686

687 - `Temperature float64`

688

689 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.

690

691 - `ToolResources AssistantToolResources`

692

693 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.

694

695 - `CodeInterpreter AssistantToolResourcesCodeInterpreter`

696

697 - `FileIDs []string`

698

699 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

700

701 - `FileSearch AssistantToolResourcesFileSearch`

702

703 - `VectorStoreIDs []string`

704

705 The ID of 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.

706

707 - `TopP float64`

708

709 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.

710

711 We generally recommend altering this or temperature but not both.

712

713### Example

714

715```go

716package main

717

718import (

719 "context"

720 "fmt"

721

722 "github.com/openai/openai-go"

723 "github.com/openai/openai-go/option"

724 "github.com/openai/openai-go/shared"

725)

726

727func main() {

728 client := openai.NewClient(

729 option.WithAPIKey("My API Key"),

730 )

731 assistant, err := client.Beta.Assistants.New(context.TODO(), openai.BetaAssistantNewParams{

732 Model: shared.ChatModelGPT4o,

733 })

734 if err != nil {

735 panic(err.Error())

736 }

737 fmt.Printf("%+v\n", assistant.ID)

738}

739```

740

741#### Response

742

743```json

744{

745 "id": "id",

746 "created_at": 0,

747 "description": "description",

748 "instructions": "instructions",

749 "metadata": {

750 "foo": "string"

751 },

752 "model": "model",

753 "name": "name",

754 "object": "assistant",

755 "tools": [

756 {}

757 ],

758 "response_format": "auto",

759 "temperature": 1,

760 "tool_resources": {

761 "code_interpreter": {

762 "file_ids": [

763 "string"

764 ]

765 },

766 "file_search": {

767 "vector_store_ids": [

768 "string"

769 ]

770 }

771 },

772 "top_p": 1

773}

774```

775

776## Retrieve assistant

777

778`client.Beta.Assistants.Get(ctx, assistantID) (*Assistant, error)`

779

780**get** `/assistants/{assistant_id}`

781

782Retrieve assistant

783

784### Parameters

785

786- `assistantID string`

787

788### Returns

789

790- `type Assistant struct{…}`

791

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

793

794 - `ID string`

795

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

797

798 - `CreatedAt int64`

799

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

801

802 - `Description string`

803

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

805

806 - `Instructions string`

807

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

809

810 - `Metadata Metadata`

811

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

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

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

815

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

817 with a maximum length of 512 characters.

818

819 - `Model string`

820

821 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

822

823 - `Name string`

824

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

826

827 - `Object Assistant`

828

829 The object type, which is always `assistant`.

830

831 - `const AssistantAssistant Assistant = "assistant"`

832

833 - `Tools []any`

834

835 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`.

836

837 - `ResponseFormat AssistantResponseFormatOptionUnion`

838

839 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`.

840

841 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).

842

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

844

845 **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.

846

847 - `type Auto string`

848

849 `auto` is the default value

850

851 - `const AutoAuto Auto = "auto"`

852

853 - `type ResponseFormatText struct{…}`

854

855 Default response format. Used to generate text responses.

856

857 - `Type Text`

858

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

860

861 - `const TextText Text = "text"`

862

863 - `type ResponseFormatJSONObject struct{…}`

864

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

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

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

868 to do so.

869

870 - `Type JSONObject`

871

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

873

874 - `const JSONObjectJSONObject JSONObject = "json_object"`

875

876 - `type ResponseFormatJSONSchema struct{…}`

877

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

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

880

881 - `JSONSchema ResponseFormatJSONSchemaJSONSchema`

882

883 Structured Outputs configuration options, including a JSON Schema.

884

885 - `Name string`

886

887 The name of the response format. Must be a-z, A-Z, 0-9, or contain

888 underscores and dashes, with a maximum length of 64.

889

890 - `Description string`

891

892 A description of what the response format is for, used by the model to

893 determine how to respond in the format.

894

895 - `Schema map[string, any]`

896

897 The schema for the response format, described as a JSON Schema object.

898 Learn how to build JSON schemas [here](https://json-schema.org/).

899

900 - `Strict bool`

901

902 Whether to enable strict schema adherence when generating the output.

903 If set to true, the model will always follow the exact schema defined

904 in the `schema` field. Only a subset of JSON Schema is supported when

905 `strict` is `true`. To learn more, read the [Structured Outputs

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

907

908 - `Type JSONSchema`

909

910 The type of response format being defined. Always `json_schema`.

911

912 - `const JSONSchemaJSONSchema JSONSchema = "json_schema"`

913

914 - `Temperature float64`

915

916 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.

917

918 - `ToolResources AssistantToolResources`

919

920 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.

921

922 - `CodeInterpreter AssistantToolResourcesCodeInterpreter`

923

924 - `FileIDs []string`

925

926 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

927

928 - `FileSearch AssistantToolResourcesFileSearch`

929

930 - `VectorStoreIDs []string`

931

932 The ID of 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.

933

934 - `TopP float64`

935

936 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.

937

938 We generally recommend altering this or temperature but not both.

939

940### Example

941

942```go

943package main

944

945import (

946 "context"

947 "fmt"

948

949 "github.com/openai/openai-go"

950 "github.com/openai/openai-go/option"

951)

952

953func main() {

954 client := openai.NewClient(

955 option.WithAPIKey("My API Key"),

956 )

957 assistant, err := client.Beta.Assistants.Get(context.TODO(), "assistant_id")

958 if err != nil {

959 panic(err.Error())

960 }

961 fmt.Printf("%+v\n", assistant.ID)

962}

963```

964

965#### Response

966

967```json

968{

969 "id": "id",

970 "created_at": 0,

971 "description": "description",

972 "instructions": "instructions",

973 "metadata": {

974 "foo": "string"

975 },

976 "model": "model",

977 "name": "name",

978 "object": "assistant",

979 "tools": [

980 {}

981 ],

982 "response_format": "auto",

983 "temperature": 1,

984 "tool_resources": {

985 "code_interpreter": {

986 "file_ids": [

987 "string"

988 ]

989 },

990 "file_search": {

991 "vector_store_ids": [

992 "string"

993 ]

994 }

995 },

996 "top_p": 1

997}

998```

999

1000## Modify assistant

1001

1002`client.Beta.Assistants.Update(ctx, assistantID, body) (*Assistant, error)`

1003

1004**post** `/assistants/{assistant_id}`

1005

1006Modify assistant

1007

1008### Parameters

1009

1010- `assistantID string`

1011

1012- `body BetaAssistantUpdateParams`

1013

1014 - `Description param.Field[string]`

1015

1016 The description of the assistant. The maximum length is 512 characters.

1017

1018 - `Instructions param.Field[string]`

1019

1020 The system instructions that the assistant uses. The maximum length is 256,000 characters.

1021

1022 - `Metadata param.Field[Metadata]`

1023

1024 Set of 16 key-value pairs that can be attached to an object. This can be

1025 useful for storing additional information about the object in a structured

1026 format, and querying for objects via API or the dashboard.

1027

1028 Keys are strings with a maximum length of 64 characters. Values are strings

1029 with a maximum length of 512 characters.

1030

1031 - `Model param.Field[BetaAssistantUpdateParamsModel]`

1032

1033 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

1034

1035 - `string`

1036

1037 - `type BetaAssistantUpdateParamsModel string`

1038

1039 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

1040

1041 - `const BetaAssistantUpdateParamsModelGPT5 BetaAssistantUpdateParamsModel = "gpt-5"`

1042

1043 - `const BetaAssistantUpdateParamsModelGPT5Mini BetaAssistantUpdateParamsModel = "gpt-5-mini"`

1044

1045 - `const BetaAssistantUpdateParamsModelGPT5Nano BetaAssistantUpdateParamsModel = "gpt-5-nano"`

1046

1047 - `const BetaAssistantUpdateParamsModelGPT5_2025_08_07 BetaAssistantUpdateParamsModel = "gpt-5-2025-08-07"`

1048

1049 - `const BetaAssistantUpdateParamsModelGPT5Mini2025_08_07 BetaAssistantUpdateParamsModel = "gpt-5-mini-2025-08-07"`

1050

1051 - `const BetaAssistantUpdateParamsModelGPT5Nano2025_08_07 BetaAssistantUpdateParamsModel = "gpt-5-nano-2025-08-07"`

1052

1053 - `const BetaAssistantUpdateParamsModelGPT4_1 BetaAssistantUpdateParamsModel = "gpt-4.1"`

1054

1055 - `const BetaAssistantUpdateParamsModelGPT4_1Mini BetaAssistantUpdateParamsModel = "gpt-4.1-mini"`

1056

1057 - `const BetaAssistantUpdateParamsModelGPT4_1Nano BetaAssistantUpdateParamsModel = "gpt-4.1-nano"`

1058

1059 - `const BetaAssistantUpdateParamsModelGPT4_1_2025_04_14 BetaAssistantUpdateParamsModel = "gpt-4.1-2025-04-14"`

1060

1061 - `const BetaAssistantUpdateParamsModelGPT4_1Mini2025_04_14 BetaAssistantUpdateParamsModel = "gpt-4.1-mini-2025-04-14"`

1062

1063 - `const BetaAssistantUpdateParamsModelGPT4_1Nano2025_04_14 BetaAssistantUpdateParamsModel = "gpt-4.1-nano-2025-04-14"`

1064

1065 - `const BetaAssistantUpdateParamsModelO3Mini BetaAssistantUpdateParamsModel = "o3-mini"`

1066

1067 - `const BetaAssistantUpdateParamsModelO3Mini2025_01_31 BetaAssistantUpdateParamsModel = "o3-mini-2025-01-31"`

1068

1069 - `const BetaAssistantUpdateParamsModelO1 BetaAssistantUpdateParamsModel = "o1"`

1070

1071 - `const BetaAssistantUpdateParamsModelO1_2024_12_17 BetaAssistantUpdateParamsModel = "o1-2024-12-17"`

1072

1073 - `const BetaAssistantUpdateParamsModelGPT4o BetaAssistantUpdateParamsModel = "gpt-4o"`

1074

1075 - `const BetaAssistantUpdateParamsModelGPT4o2024_11_20 BetaAssistantUpdateParamsModel = "gpt-4o-2024-11-20"`

1076

1077 - `const BetaAssistantUpdateParamsModelGPT4o2024_08_06 BetaAssistantUpdateParamsModel = "gpt-4o-2024-08-06"`

1078

1079 - `const BetaAssistantUpdateParamsModelGPT4o2024_05_13 BetaAssistantUpdateParamsModel = "gpt-4o-2024-05-13"`

1080

1081 - `const BetaAssistantUpdateParamsModelGPT4oMini BetaAssistantUpdateParamsModel = "gpt-4o-mini"`

1082

1083 - `const BetaAssistantUpdateParamsModelGPT4oMini2024_07_18 BetaAssistantUpdateParamsModel = "gpt-4o-mini-2024-07-18"`

1084

1085 - `const BetaAssistantUpdateParamsModelGPT4_5Preview BetaAssistantUpdateParamsModel = "gpt-4.5-preview"`

1086

1087 - `const BetaAssistantUpdateParamsModelGPT4_5Preview2025_02_27 BetaAssistantUpdateParamsModel = "gpt-4.5-preview-2025-02-27"`

1088

1089 - `const BetaAssistantUpdateParamsModelGPT4Turbo BetaAssistantUpdateParamsModel = "gpt-4-turbo"`

1090

1091 - `const BetaAssistantUpdateParamsModelGPT4Turbo2024_04_09 BetaAssistantUpdateParamsModel = "gpt-4-turbo-2024-04-09"`

1092

1093 - `const BetaAssistantUpdateParamsModelGPT4_0125Preview BetaAssistantUpdateParamsModel = "gpt-4-0125-preview"`

1094

1095 - `const BetaAssistantUpdateParamsModelGPT4TurboPreview BetaAssistantUpdateParamsModel = "gpt-4-turbo-preview"`

1096

1097 - `const BetaAssistantUpdateParamsModelGPT4_1106Preview BetaAssistantUpdateParamsModel = "gpt-4-1106-preview"`

1098

1099 - `const BetaAssistantUpdateParamsModelGPT4VisionPreview BetaAssistantUpdateParamsModel = "gpt-4-vision-preview"`

1100

1101 - `const BetaAssistantUpdateParamsModelGPT4 BetaAssistantUpdateParamsModel = "gpt-4"`

1102

1103 - `const BetaAssistantUpdateParamsModelGPT4_0314 BetaAssistantUpdateParamsModel = "gpt-4-0314"`

1104

1105 - `const BetaAssistantUpdateParamsModelGPT4_0613 BetaAssistantUpdateParamsModel = "gpt-4-0613"`

1106

1107 - `const BetaAssistantUpdateParamsModelGPT4_32k BetaAssistantUpdateParamsModel = "gpt-4-32k"`

1108

1109 - `const BetaAssistantUpdateParamsModelGPT4_32k0314 BetaAssistantUpdateParamsModel = "gpt-4-32k-0314"`

1110

1111 - `const BetaAssistantUpdateParamsModelGPT4_32k0613 BetaAssistantUpdateParamsModel = "gpt-4-32k-0613"`

1112

1113 - `const BetaAssistantUpdateParamsModelGPT3_5Turbo BetaAssistantUpdateParamsModel = "gpt-3.5-turbo"`

1114

1115 - `const BetaAssistantUpdateParamsModelGPT3_5Turbo16k BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-16k"`

1116

1117 - `const BetaAssistantUpdateParamsModelGPT3_5Turbo0613 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-0613"`

1118

1119 - `const BetaAssistantUpdateParamsModelGPT3_5Turbo1106 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-1106"`

1120

1121 - `const BetaAssistantUpdateParamsModelGPT3_5Turbo0125 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-0125"`

1122

1123 - `const BetaAssistantUpdateParamsModelGPT3_5Turbo16k0613 BetaAssistantUpdateParamsModel = "gpt-3.5-turbo-16k-0613"`

1124

1125 - `Name param.Field[string]`

1126

1127 The name of the assistant. The maximum length is 256 characters.

1128

1129 - `ReasoningEffort param.Field[ReasoningEffort]`

1130

1131 Constrains effort on reasoning for reasoning models. Currently supported

1132 values are `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, and `max`.

1133 Reducing reasoning effort can result in faster responses and fewer tokens

1134 used on reasoning in a response. Not all reasoning models support every

1135 value. See the

1136 [reasoning guide](https://platform.openai.com/docs/guides/reasoning)

1137 for model-specific support.

1138

1139 - `ResponseFormat param.Field[AssistantResponseFormatOptionUnion]`

1140

1141 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`.

1142

1143 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).

1144

1145 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

1146

1147 **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.

1148

1149 - `Temperature param.Field[float64]`

1150

1151 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.

1152

1153 - `ToolResources param.Field[BetaAssistantUpdateParamsToolResources]`

1154

1155 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.

1156

1157 - `CodeInterpreter BetaAssistantUpdateParamsToolResourcesCodeInterpreter`

1158

1159 - `FileIDs []string`

1160

1161 Overrides the list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter` tool. There can be a maximum of 20 files associated with the tool.

1162

1163 - `FileSearch BetaAssistantUpdateParamsToolResourcesFileSearch`

1164

1165 - `VectorStoreIDs []string`

1166

1167 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.

1168

1169 - `Tools param.Field[[]any]`

1170

1171 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`.

1172

1173 - `TopP param.Field[float64]`

1174

1175 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.

1176

1177 We generally recommend altering this or temperature but not both.

1178

1179### Returns

1180

1181- `type Assistant struct{…}`

1182

1183 Represents an `assistant` that can call the model and use tools.

1184

1185 - `ID string`

1186

1187 The identifier, which can be referenced in API endpoints.

1188

1189 - `CreatedAt int64`

1190

1191 The Unix timestamp (in seconds) for when the assistant was created.

1192

1193 - `Description string`

1194

1195 The description of the assistant. The maximum length is 512 characters.

1196

1197 - `Instructions string`

1198

1199 The system instructions that the assistant uses. The maximum length is 256,000 characters.

1200

1201 - `Metadata Metadata`

1202

1203 Set of 16 key-value pairs that can be attached to an object. This can be

1204 useful for storing additional information about the object in a structured

1205 format, and querying for objects via API or the dashboard.

1206

1207 Keys are strings with a maximum length of 64 characters. Values are strings

1208 with a maximum length of 512 characters.

1209

1210 - `Model string`

1211

1212 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

1213

1214 - `Name string`

1215

1216 The name of the assistant. The maximum length is 256 characters.

1217

1218 - `Object Assistant`

1219

1220 The object type, which is always `assistant`.

1221

1222 - `const AssistantAssistant Assistant = "assistant"`

1223

1224 - `Tools []any`

1225

1226 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`.

1227

1228 - `ResponseFormat AssistantResponseFormatOptionUnion`

1229

1230 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`.

1231

1232 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).

1233

1234 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

1235

1236 **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.

1237

1238 - `type Auto string`

1239

1240 `auto` is the default value

1241

1242 - `const AutoAuto Auto = "auto"`

1243

1244 - `type ResponseFormatText struct{…}`

1245

1246 Default response format. Used to generate text responses.

1247

1248 - `Type Text`

1249

1250 The type of response format being defined. Always `text`.

1251

1252 - `const TextText Text = "text"`

1253

1254 - `type ResponseFormatJSONObject struct{…}`

1255

1256 JSON object response format. An older method of generating JSON responses.

1257 Using `json_schema` is recommended for models that support it. Note that the

1258 model will not generate JSON without a system or user message instructing it

1259 to do so.

1260

1261 - `Type JSONObject`

1262

1263 The type of response format being defined. Always `json_object`.

1264

1265 - `const JSONObjectJSONObject JSONObject = "json_object"`

1266

1267 - `type ResponseFormatJSONSchema struct{…}`

1268

1269 JSON Schema response format. Used to generate structured JSON responses.

1270 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).

1271

1272 - `JSONSchema ResponseFormatJSONSchemaJSONSchema`

1273

1274 Structured Outputs configuration options, including a JSON Schema.

1275

1276 - `Name string`

1277

1278 The name of the response format. Must be a-z, A-Z, 0-9, or contain

1279 underscores and dashes, with a maximum length of 64.

1280

1281 - `Description string`

1282

1283 A description of what the response format is for, used by the model to

1284 determine how to respond in the format.

1285

1286 - `Schema map[string, any]`

1287

1288 The schema for the response format, described as a JSON Schema object.

1289 Learn how to build JSON schemas [here](https://json-schema.org/).

1290

1291 - `Strict bool`

1292

1293 Whether to enable strict schema adherence when generating the output.

1294 If set to true, the model will always follow the exact schema defined

1295 in the `schema` field. Only a subset of JSON Schema is supported when

1296 `strict` is `true`. To learn more, read the [Structured Outputs

1297 guide](https://platform.openai.com/docs/guides/structured-outputs).

1298

1299 - `Type JSONSchema`

1300

1301 The type of response format being defined. Always `json_schema`.

1302

1303 - `const JSONSchemaJSONSchema JSONSchema = "json_schema"`

1304

1305 - `Temperature float64`

1306

1307 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.

1308

1309 - `ToolResources AssistantToolResources`

1310

1311 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.

1312

1313 - `CodeInterpreter AssistantToolResourcesCodeInterpreter`

1314

1315 - `FileIDs []string`

1316

1317 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

1318

1319 - `FileSearch AssistantToolResourcesFileSearch`

1320

1321 - `VectorStoreIDs []string`

1322

1323 The ID of 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.

1324

1325 - `TopP float64`

1326

1327 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.

1328

1329 We generally recommend altering this or temperature but not both.

1330

1331### Example

1332

1333```go

1334package main

1335

1336import (

1337 "context"

1338 "fmt"

1339

1340 "github.com/openai/openai-go"

1341 "github.com/openai/openai-go/option"

1342)

1343

1344func main() {

1345 client := openai.NewClient(

1346 option.WithAPIKey("My API Key"),

1347 )

1348 assistant, err := client.Beta.Assistants.Update(

1349 context.TODO(),

1350 "assistant_id",

1351 openai.BetaAssistantUpdateParams{

1352

1353 },

1354 )

1355 if err != nil {

1356 panic(err.Error())

1357 }

1358 fmt.Printf("%+v\n", assistant.ID)

1359}

1360```

1361

1362#### Response

1363

1364```json

1365{

1366 "id": "id",

1367 "created_at": 0,

1368 "description": "description",

1369 "instructions": "instructions",

1370 "metadata": {

1371 "foo": "string"

1372 },

1373 "model": "model",

1374 "name": "name",

1375 "object": "assistant",

1376 "tools": [

1377 {}

1378 ],

1379 "response_format": "auto",

1380 "temperature": 1,

1381 "tool_resources": {

1382 "code_interpreter": {

1383 "file_ids": [

1384 "string"

1385 ]

1386 },

1387 "file_search": {

1388 "vector_store_ids": [

1389 "string"

1390 ]

1391 }

1392 },

1393 "top_p": 1

1394}

1395```

1396

1397## Delete assistant

1398

1399`client.Beta.Assistants.Delete(ctx, assistantID) (*AssistantDeleted, error)`

1400

1401**delete** `/assistants/{assistant_id}`

1402

1403Delete assistant

1404

1405### Parameters

1406

1407- `assistantID string`

1408

1409### Returns

1410

1411- `type AssistantDeleted struct{…}`

1412

1413 - `ID string`

1414

1415 - `Deleted bool`

1416

1417 - `Object AssistantDeleted`

1418

1419 - `const AssistantDeletedAssistantDeleted AssistantDeleted = "assistant.deleted"`

1420

1421### Example

1422

1423```go

1424package main

1425

1426import (

1427 "context"

1428 "fmt"

1429

1430 "github.com/openai/openai-go"

1431 "github.com/openai/openai-go/option"

1432)

1433

1434func main() {

1435 client := openai.NewClient(

1436 option.WithAPIKey("My API Key"),

1437 )

1438 assistantDeleted, err := client.Beta.Assistants.Delete(context.TODO(), "assistant_id")

1439 if err != nil {

1440 panic(err.Error())

1441 }

1442 fmt.Printf("%+v\n", assistantDeleted.ID)

1443}

1444```

1445

1446#### Response

1447

1448```json

1449{

1450 "id": "id",

1451 "deleted": true,

1452 "object": "assistant.deleted"

1453}

1454```

1455

1456## Domain Types

1457

1458### Assistant

1459

1460- `type Assistant struct{…}`

1461

1462 Represents an `assistant` that can call the model and use tools.

1463

1464 - `ID string`

1465

1466 The identifier, which can be referenced in API endpoints.

1467

1468 - `CreatedAt int64`

1469

1470 The Unix timestamp (in seconds) for when the assistant was created.

1471

1472 - `Description string`

1473

1474 The description of the assistant. The maximum length is 512 characters.

1475

1476 - `Instructions string`

1477

1478 The system instructions that the assistant uses. The maximum length is 256,000 characters.

1479

1480 - `Metadata Metadata`

1481

1482 Set of 16 key-value pairs that can be attached to an object. This can be

1483 useful for storing additional information about the object in a structured

1484 format, and querying for objects via API or the dashboard.

1485

1486 Keys are strings with a maximum length of 64 characters. Values are strings

1487 with a maximum length of 512 characters.

1488

1489 - `Model string`

1490

1491 ID of the model to use. You can use the [List models](https://platform.openai.com/docs/api-reference/models/list) API to see all of your available models, or see our [Model overview](https://platform.openai.com/docs/models) for descriptions of them.

1492

1493 - `Name string`

1494

1495 The name of the assistant. The maximum length is 256 characters.

1496

1497 - `Object Assistant`

1498

1499 The object type, which is always `assistant`.

1500

1501 - `const AssistantAssistant Assistant = "assistant"`

1502

1503 - `Tools []any`

1504

1505 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`.

1506

1507 - `ResponseFormat AssistantResponseFormatOptionUnion`

1508

1509 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`.

1510

1511 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).

1512

1513 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

1514

1515 **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.

1516

1517 - `type Auto string`

1518

1519 `auto` is the default value

1520

1521 - `const AutoAuto Auto = "auto"`

1522

1523 - `type ResponseFormatText struct{…}`

1524

1525 Default response format. Used to generate text responses.

1526

1527 - `Type Text`

1528

1529 The type of response format being defined. Always `text`.

1530

1531 - `const TextText Text = "text"`

1532

1533 - `type ResponseFormatJSONObject struct{…}`

1534

1535 JSON object response format. An older method of generating JSON responses.

1536 Using `json_schema` is recommended for models that support it. Note that the

1537 model will not generate JSON without a system or user message instructing it

1538 to do so.

1539

1540 - `Type JSONObject`

1541

1542 The type of response format being defined. Always `json_object`.

1543

1544 - `const JSONObjectJSONObject JSONObject = "json_object"`

1545

1546 - `type ResponseFormatJSONSchema struct{…}`

1547

1548 JSON Schema response format. Used to generate structured JSON responses.

1549 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).

1550

1551 - `JSONSchema ResponseFormatJSONSchemaJSONSchema`

1552

1553 Structured Outputs configuration options, including a JSON Schema.

1554

1555 - `Name string`

1556

1557 The name of the response format. Must be a-z, A-Z, 0-9, or contain

1558 underscores and dashes, with a maximum length of 64.

1559

1560 - `Description string`

1561

1562 A description of what the response format is for, used by the model to

1563 determine how to respond in the format.

1564

1565 - `Schema map[string, any]`

1566

1567 The schema for the response format, described as a JSON Schema object.

1568 Learn how to build JSON schemas [here](https://json-schema.org/).

1569

1570 - `Strict bool`

1571

1572 Whether to enable strict schema adherence when generating the output.

1573 If set to true, the model will always follow the exact schema defined

1574 in the `schema` field. Only a subset of JSON Schema is supported when

1575 `strict` is `true`. To learn more, read the [Structured Outputs

1576 guide](https://platform.openai.com/docs/guides/structured-outputs).

1577

1578 - `Type JSONSchema`

1579

1580 The type of response format being defined. Always `json_schema`.

1581

1582 - `const JSONSchemaJSONSchema JSONSchema = "json_schema"`

1583

1584 - `Temperature float64`

1585

1586 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.

1587

1588 - `ToolResources AssistantToolResources`

1589

1590 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.

1591

1592 - `CodeInterpreter AssistantToolResourcesCodeInterpreter`

1593

1594 - `FileIDs []string`

1595

1596 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter`` tool. There can be a maximum of 20 files associated with the tool.

1597

1598 - `FileSearch AssistantToolResourcesFileSearch`

1599

1600 - `VectorStoreIDs []string`

1601

1602 The ID of 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.

1603

1604 - `TopP float64`

1605

1606 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.

1607

1608 We generally recommend altering this or temperature but not both.

1609

1610### Assistant Deleted

1611

1612- `type AssistantDeleted struct{…}`

1613

1614 - `ID string`

1615

1616 - `Deleted bool`

1617

1618 - `Object AssistantDeleted`

1619

1620 - `const AssistantDeletedAssistantDeleted AssistantDeleted = "assistant.deleted"`

1621

1622### Assistant Stream Event

1623

1624- `type AssistantStreamEventUnion interface{…}`

1625

1626 Represents an event emitted when streaming a Run.

1627

1628 Each event in a server-sent events stream has an `event` and `data` property:

1629

1630 ```

1631 event: thread.created

1632 data: {"id": "thread_123", "object": "thread", ...}

1633 ```

1634

1635 We emit events whenever a new object is created, transitions to a new state, or is being

1636 streamed in parts (deltas). For example, we emit `thread.run.created` when a new run

1637 is created, `thread.run.completed` when a run completes, and so on. When an Assistant chooses

1638 to create a message during a run, we emit a `thread.message.created event`, a

1639 `thread.message.in_progress` event, many `thread.message.delta` events, and finally a

1640 `thread.message.completed` event.

1641

1642 We may add additional events over time, so we recommend handling unknown events gracefully

1643 in your code. See the [Assistants API quickstart](https://platform.openai.com/docs/assistants/overview) to learn how to

1644 integrate the Assistants API with streaming.

1645

1646 - `type AssistantStreamEventThreadCreated struct{…}`

1647

1648 Occurs when a new [thread](https://platform.openai.com/docs/api-reference/threads/object) is created.

1649

1650 - `Data Thread`

1651

1652 Represents a thread that contains [messages](https://platform.openai.com/docs/api-reference/messages).

1653

1654 - `ID string`

1655

1656 The identifier, which can be referenced in API endpoints.

1657

1658 - `CreatedAt int64`

1659

1660 The Unix timestamp (in seconds) for when the thread was created.

1661

1662 - `Metadata Metadata`

1663

1664 Set of 16 key-value pairs that can be attached to an object. This can be

1665 useful for storing additional information about the object in a structured

1666 format, and querying for objects via API or the dashboard.

1667

1668 Keys are strings with a maximum length of 64 characters. Values are strings

1669 with a maximum length of 512 characters.

1670

1671 - `Object Thread`

1672

1673 The object type, which is always `thread`.

1674

1675 - `const ThreadThread Thread = "thread"`

1676

1677 - `ToolResources ThreadToolResources`

1678

1679 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.

1680

1681 - `CodeInterpreter ThreadToolResourcesCodeInterpreter`

1682

1683 - `FileIDs []string`

1684

1685 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter` tool. There can be a maximum of 20 files associated with the tool.

1686

1687 - `FileSearch ThreadToolResourcesFileSearch`

1688

1689 - `VectorStoreIDs []string`

1690

1691 The [vector store](https://platform.openai.com/docs/api-reference/vector-stores/object) attached to this thread. There can be a maximum of 1 vector store attached to the thread.

1692

1693 - `Event ThreadCreated`

1694

1695 - `const ThreadCreatedThreadCreated ThreadCreated = "thread.created"`

1696

1697 - `Enabled bool`

1698

1699 Whether to enable input audio transcription.

1700

1701 - `type AssistantStreamEventThreadRunCreated struct{…}`

1702

1703 Occurs when a new [run](https://platform.openai.com/docs/api-reference/runs/object) is created.

1704

1705 - `Data Run`

1706

1707 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

1708

1709 - `ID string`

1710

1711 The identifier, which can be referenced in API endpoints.

1712

1713 - `AssistantID string`

1714

1715 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for execution of this run.

1716

1717 - `CancelledAt int64`

1718

1719 The Unix timestamp (in seconds) for when the run was cancelled.

1720

1721 - `CompletedAt int64`

1722

1723 The Unix timestamp (in seconds) for when the run was completed.

1724

1725 - `CreatedAt int64`

1726

1727 The Unix timestamp (in seconds) for when the run was created.

1728

1729 - `ExpiresAt int64`

1730

1731 The Unix timestamp (in seconds) for when the run will expire.

1732

1733 - `FailedAt int64`

1734

1735 The Unix timestamp (in seconds) for when the run failed.

1736

1737 - `IncompleteDetails RunIncompleteDetails`

1738

1739 Details on why the run is incomplete. Will be `null` if the run is not incomplete.

1740

1741 - `Reason string`

1742

1743 The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.

1744

1745 - `const RunIncompleteDetailsReasonMaxCompletionTokens RunIncompleteDetailsReason = "max_completion_tokens"`

1746

1747 - `const RunIncompleteDetailsReasonMaxPromptTokens RunIncompleteDetailsReason = "max_prompt_tokens"`

1748

1749 - `Instructions string`

1750

1751 The instructions that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

1752

1753 - `LastError RunLastError`

1754

1755 The last error associated with this run. Will be `null` if there are no errors.

1756

1757 - `Code string`

1758

1759 One of `server_error`, `rate_limit_exceeded`, or `invalid_prompt`.

1760

1761 - `const RunLastErrorCodeServerError RunLastErrorCode = "server_error"`

1762

1763 - `const RunLastErrorCodeRateLimitExceeded RunLastErrorCode = "rate_limit_exceeded"`

1764

1765 - `const RunLastErrorCodeInvalidPrompt RunLastErrorCode = "invalid_prompt"`

1766

1767 - `Message string`

1768

1769 A human-readable description of the error.

1770

1771 - `MaxCompletionTokens int64`

1772

1773 The maximum number of completion tokens specified to have been used over the course of the run.

1774

1775 - `MaxPromptTokens int64`

1776

1777 The maximum number of prompt tokens specified to have been used over the course of the run.

1778

1779 - `Metadata Metadata`

1780

1781 Set of 16 key-value pairs that can be attached to an object. This can be

1782 useful for storing additional information about the object in a structured

1783 format, and querying for objects via API or the dashboard.

1784

1785 Keys are strings with a maximum length of 64 characters. Values are strings

1786 with a maximum length of 512 characters.

1787

1788 - `Model string`

1789

1790 The model that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

1791

1792 - `Object ThreadRun`

1793

1794 The object type, which is always `thread.run`.

1795

1796 - `const ThreadRunThreadRun ThreadRun = "thread.run"`

1797

1798 - `ParallelToolCalls bool`

1799

1800 Whether to enable [parallel function calling](https://platform.openai.com/docs/guides/function-calling#configuring-parallel-function-calling) during tool use.

1801

1802 - `RequiredAction RunRequiredAction`

1803

1804 Details on the action required to continue the run. Will be `null` if no action is required.

1805

1806 - `SubmitToolOutputs RunRequiredActionSubmitToolOutputs`

1807

1808 Details on the tool outputs needed for this run to continue.

1809

1810 - `ToolCalls []RequiredActionFunctionToolCall`

1811

1812 A list of the relevant tool calls.

1813

1814 - `ID string`

1815

1816 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](https://platform.openai.com/docs/api-reference/runs/submitToolOutputs) endpoint.

1817

1818 - `Function RequiredActionFunctionToolCallFunction`

1819

1820 The function definition.

1821

1822 - `Arguments string`

1823

1824 The arguments that the model expects you to pass to the function.

1825

1826 - `Name string`

1827

1828 The name of the function.

1829

1830 - `Type Function`

1831

1832 The type of tool call the output is required for. For now, this is always `function`.

1833

1834 - `const FunctionFunction Function = "function"`

1835

1836 - `Type SubmitToolOutputs`

1837

1838 For now, this is always `submit_tool_outputs`.

1839

1840 - `const SubmitToolOutputsSubmitToolOutputs SubmitToolOutputs = "submit_tool_outputs"`

1841

1842 - `ResponseFormat AssistantResponseFormatOptionUnion`

1843

1844 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`.

1845

1846 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).

1847

1848 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

1849

1850 **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.

1851

1852 - `type Auto string`

1853

1854 `auto` is the default value

1855

1856 - `const AutoAuto Auto = "auto"`

1857

1858 - `type ResponseFormatText struct{…}`

1859

1860 Default response format. Used to generate text responses.

1861

1862 - `Type Text`

1863

1864 The type of response format being defined. Always `text`.

1865

1866 - `const TextText Text = "text"`

1867

1868 - `type ResponseFormatJSONObject struct{…}`

1869

1870 JSON object response format. An older method of generating JSON responses.

1871 Using `json_schema` is recommended for models that support it. Note that the

1872 model will not generate JSON without a system or user message instructing it

1873 to do so.

1874

1875 - `Type JSONObject`

1876

1877 The type of response format being defined. Always `json_object`.

1878

1879 - `const JSONObjectJSONObject JSONObject = "json_object"`

1880

1881 - `type ResponseFormatJSONSchema struct{…}`

1882

1883 JSON Schema response format. Used to generate structured JSON responses.

1884 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).

1885

1886 - `JSONSchema ResponseFormatJSONSchemaJSONSchema`

1887

1888 Structured Outputs configuration options, including a JSON Schema.

1889

1890 - `Name string`

1891

1892 The name of the response format. Must be a-z, A-Z, 0-9, or contain

1893 underscores and dashes, with a maximum length of 64.

1894

1895 - `Description string`

1896

1897 A description of what the response format is for, used by the model to

1898 determine how to respond in the format.

1899

1900 - `Schema map[string, any]`

1901

1902 The schema for the response format, described as a JSON Schema object.

1903 Learn how to build JSON schemas [here](https://json-schema.org/).

1904

1905 - `Strict bool`

1906

1907 Whether to enable strict schema adherence when generating the output.

1908 If set to true, the model will always follow the exact schema defined

1909 in the `schema` field. Only a subset of JSON Schema is supported when

1910 `strict` is `true`. To learn more, read the [Structured Outputs

1911 guide](https://platform.openai.com/docs/guides/structured-outputs).

1912

1913 - `Type JSONSchema`

1914

1915 The type of response format being defined. Always `json_schema`.

1916

1917 - `const JSONSchemaJSONSchema JSONSchema = "json_schema"`

1918

1919 - `StartedAt int64`

1920

1921 The Unix timestamp (in seconds) for when the run was started.

1922

1923 - `Status any`

1924

1925 - `ThreadID string`

1926

1927 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was executed on as a part of this run.

1928

1929 - `ToolChoice AssistantToolChoiceOptionUnion`

1930

1931 Controls which (if any) tool is called by the model.

1932 `none` means the model will not call any tools and instead generates a message.

1933 `auto` is the default value and means the model can pick between generating a message or calling one or more tools.

1934 `required` means the model must call one or more tools before responding to the user.

1935 Specifying a particular tool like `{"type": "file_search"}` or `{"type": "function", "function": {"name": "my_function"}}` forces the model to call that tool.

1936

1937 - `type AssistantToolChoiceOptionAuto string`

1938

1939 `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.

1940

1941 - `const AssistantToolChoiceOptionAutoNone AssistantToolChoiceOptionAuto = "none"`

1942

1943 - `const AssistantToolChoiceOptionAutoAuto AssistantToolChoiceOptionAuto = "auto"`

1944

1945 - `const AssistantToolChoiceOptionAutoRequired AssistantToolChoiceOptionAuto = "required"`

1946

1947 - `type AssistantToolChoice struct{…}`

1948

1949 Specifies a tool the model should use. Use to force the model to call a specific tool.

1950

1951 - `Type AssistantToolChoiceType`

1952

1953 The type of the tool. If type is `function`, the function name must be set

1954

1955 - `const AssistantToolChoiceTypeFunction AssistantToolChoiceType = "function"`

1956

1957 - `const AssistantToolChoiceTypeCodeInterpreter AssistantToolChoiceType = "code_interpreter"`

1958

1959 - `const AssistantToolChoiceTypeFileSearch AssistantToolChoiceType = "file_search"`

1960

1961 - `Function AssistantToolChoiceFunction`

1962

1963 - `Name string`

1964

1965 The name of the function to call.

1966

1967 - `Tools []any`

1968

1969 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

1970

1971 - `TruncationStrategy RunTruncationStrategy`

1972

1973 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

1974

1975 - `Type string`

1976

1977 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`.

1978

1979 - `const RunTruncationStrategyTypeAuto RunTruncationStrategyType = "auto"`

1980

1981 - `const RunTruncationStrategyTypeLastMessages RunTruncationStrategyType = "last_messages"`

1982

1983 - `LastMessages int64`

1984

1985 The number of most recent messages from the thread when constructing the context for the run.

1986

1987 - `Usage RunUsage`

1988

1989 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.).

1990

1991 - `CompletionTokens int64`

1992

1993 Number of completion tokens used over the course of the run.

1994

1995 - `PromptTokens int64`

1996

1997 Number of prompt tokens used over the course of the run.

1998

1999 - `TotalTokens int64`

2000

2001 Total number of tokens used (prompt + completion).

2002

2003 - `Temperature float64`

2004

2005 The sampling temperature used for this run. If not set, defaults to 1.

2006

2007 - `TopP float64`

2008

2009 The nucleus sampling value used for this run. If not set, defaults to 1.

2010

2011 - `Event ThreadRunCreated`

2012

2013 - `const ThreadRunCreatedThreadRunCreated ThreadRunCreated = "thread.run.created"`

2014

2015 - `type AssistantStreamEventThreadRunQueued struct{…}`

2016

2017 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) moves to a `queued` status.

2018

2019 - `Data Run`

2020

2021 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2022

2023 - `Event ThreadRunQueued`

2024

2025 - `const ThreadRunQueuedThreadRunQueued ThreadRunQueued = "thread.run.queued"`

2026

2027 - `type AssistantStreamEventThreadRunInProgress struct{…}`

2028

2029 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) moves to an `in_progress` status.

2030

2031 - `Data Run`

2032

2033 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2034

2035 - `Event ThreadRunInProgress`

2036

2037 - `const ThreadRunInProgressThreadRunInProgress ThreadRunInProgress = "thread.run.in_progress"`

2038

2039 - `type AssistantStreamEventThreadRunRequiresAction struct{…}`

2040

2041 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) moves to a `requires_action` status.

2042

2043 - `Data Run`

2044

2045 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2046

2047 - `Event ThreadRunRequiresAction`

2048

2049 - `const ThreadRunRequiresActionThreadRunRequiresAction ThreadRunRequiresAction = "thread.run.requires_action"`

2050

2051 - `type AssistantStreamEventThreadRunCompleted struct{…}`

2052

2053 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) is completed.

2054

2055 - `Data Run`

2056

2057 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2058

2059 - `Event ThreadRunCompleted`

2060

2061 - `const ThreadRunCompletedThreadRunCompleted ThreadRunCompleted = "thread.run.completed"`

2062

2063 - `type AssistantStreamEventThreadRunIncomplete struct{…}`

2064

2065 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) ends with status `incomplete`.

2066

2067 - `Data Run`

2068

2069 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2070

2071 - `Event ThreadRunIncomplete`

2072

2073 - `const ThreadRunIncompleteThreadRunIncomplete ThreadRunIncomplete = "thread.run.incomplete"`

2074

2075 - `type AssistantStreamEventThreadRunFailed struct{…}`

2076

2077 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) fails.

2078

2079 - `Data Run`

2080

2081 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2082

2083 - `Event ThreadRunFailed`

2084

2085 - `const ThreadRunFailedThreadRunFailed ThreadRunFailed = "thread.run.failed"`

2086

2087 - `type AssistantStreamEventThreadRunCancelling struct{…}`

2088

2089 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) moves to a `cancelling` status.

2090

2091 - `Data Run`

2092

2093 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2094

2095 - `Event ThreadRunCancelling`

2096

2097 - `const ThreadRunCancellingThreadRunCancelling ThreadRunCancelling = "thread.run.cancelling"`

2098

2099 - `type AssistantStreamEventThreadRunCancelled struct{…}`

2100

2101 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) is cancelled.

2102

2103 - `Data Run`

2104

2105 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2106

2107 - `Event ThreadRunCancelled`

2108

2109 - `const ThreadRunCancelledThreadRunCancelled ThreadRunCancelled = "thread.run.cancelled"`

2110

2111 - `type AssistantStreamEventThreadRunExpired struct{…}`

2112

2113 Occurs when a [run](https://platform.openai.com/docs/api-reference/runs/object) expires.

2114

2115 - `Data Run`

2116

2117 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

2118

2119 - `Event ThreadRunExpired`

2120

2121 - `const ThreadRunExpiredThreadRunExpired ThreadRunExpired = "thread.run.expired"`

2122

2123 - `type AssistantStreamEventThreadRunStepCreated struct{…}`

2124

2125 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.

2126

2127 - `Data RunStep`

2128

2129 Represents a step in execution of a run.

2130

2131 - `ID string`

2132

2133 The identifier of the run step, which can be referenced in API endpoints.

2134

2135 - `AssistantID string`

2136

2137 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) associated with the run step.

2138

2139 - `CancelledAt int64`

2140

2141 The Unix timestamp (in seconds) for when the run step was cancelled.

2142

2143 - `CompletedAt int64`

2144

2145 The Unix timestamp (in seconds) for when the run step completed.

2146

2147 - `CreatedAt int64`

2148

2149 The Unix timestamp (in seconds) for when the run step was created.

2150

2151 - `ExpiredAt int64`

2152

2153 The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.

2154

2155 - `FailedAt int64`

2156

2157 The Unix timestamp (in seconds) for when the run step failed.

2158

2159 - `LastError RunStepLastError`

2160

2161 The last error associated with this run step. Will be `null` if there are no errors.

2162

2163 - `Code string`

2164

2165 One of `server_error` or `rate_limit_exceeded`.

2166

2167 - `const RunStepLastErrorCodeServerError RunStepLastErrorCode = "server_error"`

2168

2169 - `const RunStepLastErrorCodeRateLimitExceeded RunStepLastErrorCode = "rate_limit_exceeded"`

2170

2171 - `Message string`

2172

2173 A human-readable description of the error.

2174

2175 - `Metadata Metadata`

2176

2177 Set of 16 key-value pairs that can be attached to an object. This can be

2178 useful for storing additional information about the object in a structured

2179 format, and querying for objects via API or the dashboard.

2180

2181 Keys are strings with a maximum length of 64 characters. Values are strings

2182 with a maximum length of 512 characters.

2183

2184 - `Object ThreadRunStep`

2185

2186 The object type, which is always `thread.run.step`.

2187

2188 - `const ThreadRunStepThreadRunStep ThreadRunStep = "thread.run.step"`

2189

2190 - `RunID string`

2191

2192 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) that this run step is a part of.

2193

2194 - `Status RunStepStatus`

2195

2196 The status of the run step, which can be either `in_progress`, `cancelled`, `failed`, `completed`, or `expired`.

2197

2198 - `const RunStepStatusInProgress RunStepStatus = "in_progress"`

2199

2200 - `const RunStepStatusCancelled RunStepStatus = "cancelled"`

2201

2202 - `const RunStepStatusFailed RunStepStatus = "failed"`

2203

2204 - `const RunStepStatusCompleted RunStepStatus = "completed"`

2205

2206 - `const RunStepStatusExpired RunStepStatus = "expired"`

2207

2208 - `StepDetails RunStepStepDetailsUnion`

2209

2210 The details of the run step.

2211

2212 - `type MessageCreationStepDetails struct{…}`

2213

2214 Details of the message creation by the run step.

2215

2216 - `MessageCreation MessageCreationStepDetailsMessageCreation`

2217

2218 - `MessageID string`

2219

2220 The ID of the message that was created by this run step.

2221

2222 - `Type MessageCreation`

2223

2224 Always `message_creation`.

2225

2226 - `const MessageCreationMessageCreation MessageCreation = "message_creation"`

2227

2228 - `type ToolCallsStepDetails struct{…}`

2229

2230 Details of the tool call.

2231

2232 - `ToolCalls []any`

2233

2234 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`.

2235

2236 - `Type ToolCalls`

2237

2238 Always `tool_calls`.

2239

2240 - `const ToolCallsToolCalls ToolCalls = "tool_calls"`

2241

2242 - `ThreadID string`

2243

2244 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.

2245

2246 - `Type RunStepType`

2247

2248 The type of run step, which can be either `message_creation` or `tool_calls`.

2249

2250 - `const RunStepTypeMessageCreation RunStepType = "message_creation"`

2251

2252 - `const RunStepTypeToolCalls RunStepType = "tool_calls"`

2253

2254 - `Usage RunStepUsage`

2255

2256 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.

2257

2258 - `CompletionTokens int64`

2259

2260 Number of completion tokens used over the course of the run step.

2261

2262 - `PromptTokens int64`

2263

2264 Number of prompt tokens used over the course of the run step.

2265

2266 - `TotalTokens int64`

2267

2268 Total number of tokens used (prompt + completion).

2269

2270 - `Event ThreadRunStepCreated`

2271

2272 - `const ThreadRunStepCreatedThreadRunStepCreated ThreadRunStepCreated = "thread.run.step.created"`

2273

2274 - `type AssistantStreamEventThreadRunStepInProgress struct{…}`

2275

2276 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) moves to an `in_progress` state.

2277

2278 - `Data RunStep`

2279

2280 Represents a step in execution of a run.

2281

2282 - `Event ThreadRunStepInProgress`

2283

2284 - `const ThreadRunStepInProgressThreadRunStepInProgress ThreadRunStepInProgress = "thread.run.step.in_progress"`

2285

2286 - `type AssistantStreamEventThreadRunStepDelta struct{…}`

2287

2288 Occurs when parts of a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) are being streamed.

2289

2290 - `Data RunStepDeltaEvent`

2291

2292 Represents a run step delta i.e. any changed fields on a run step during streaming.

2293

2294 - `ID string`

2295

2296 The identifier of the run step, which can be referenced in API endpoints.

2297

2298 - `Delta any`

2299

2300 - `Object ThreadRunStepDelta`

2301

2302 The object type, which is always `thread.run.step.delta`.

2303

2304 - `const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"`

2305

2306 - `Event ThreadRunStepDelta`

2307

2308 - `const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"`

2309

2310 - `type AssistantStreamEventThreadRunStepCompleted struct{…}`

2311

2312 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is completed.

2313

2314 - `Data RunStep`

2315

2316 Represents a step in execution of a run.

2317

2318 - `Event ThreadRunStepCompleted`

2319

2320 - `const ThreadRunStepCompletedThreadRunStepCompleted ThreadRunStepCompleted = "thread.run.step.completed"`

2321

2322 - `type AssistantStreamEventThreadRunStepFailed struct{…}`

2323

2324 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) fails.

2325

2326 - `Data RunStep`

2327

2328 Represents a step in execution of a run.

2329

2330 - `Event ThreadRunStepFailed`

2331

2332 - `const ThreadRunStepFailedThreadRunStepFailed ThreadRunStepFailed = "thread.run.step.failed"`

2333

2334 - `type AssistantStreamEventThreadRunStepCancelled struct{…}`

2335

2336 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is cancelled.

2337

2338 - `Data RunStep`

2339

2340 Represents a step in execution of a run.

2341

2342 - `Event ThreadRunStepCancelled`

2343

2344 - `const ThreadRunStepCancelledThreadRunStepCancelled ThreadRunStepCancelled = "thread.run.step.cancelled"`

2345

2346 - `type AssistantStreamEventThreadRunStepExpired struct{…}`

2347

2348 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) expires.

2349

2350 - `Data RunStep`

2351

2352 Represents a step in execution of a run.

2353

2354 - `Event ThreadRunStepExpired`

2355

2356 - `const ThreadRunStepExpiredThreadRunStepExpired ThreadRunStepExpired = "thread.run.step.expired"`

2357

2358 - `type AssistantStreamEventThreadMessageCreated struct{…}`

2359

2360 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.

2361

2362 - `Data Message`

2363

2364 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2365

2366 - `ID string`

2367

2368 The identifier, which can be referenced in API endpoints.

2369

2370 - `AssistantID string`

2371

2372 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.

2373

2374 - `Attachments []MessageAttachment`

2375

2376 A list of files attached to the message, and the tools they were added to.

2377

2378 - `FileID string`

2379

2380 The ID of the file to attach to the message.

2381

2382 - `Tools []MessageAttachmentToolUnion`

2383

2384 The tools to add this file to.

2385

2386 - `type CodeInterpreterTool struct{…}`

2387

2388 - `Type CodeInterpreter`

2389

2390 The type of tool being defined: `code_interpreter`

2391

2392 - `const CodeInterpreterCodeInterpreter CodeInterpreter = "code_interpreter"`

2393

2394 - `type MessageAttachmentToolAssistantToolsFileSearchTypeOnly struct{…}`

2395

2396 - `Type FileSearch`

2397

2398 The type of tool being defined: `file_search`

2399

2400 - `const FileSearchFileSearch FileSearch = "file_search"`

2401

2402 - `CompletedAt int64`

2403

2404 The Unix timestamp (in seconds) for when the message was completed.

2405

2406 - `Content []any`

2407

2408 The content of the message in array of text and/or images.

2409

2410 - `CreatedAt int64`

2411

2412 The Unix timestamp (in seconds) for when the message was created.

2413

2414 - `IncompleteAt int64`

2415

2416 The Unix timestamp (in seconds) for when the message was marked as incomplete.

2417

2418 - `IncompleteDetails MessageIncompleteDetails`

2419

2420 On an incomplete message, details about why the message is incomplete.

2421

2422 - `Reason string`

2423

2424 The reason the message is incomplete.

2425

2426 - `const MessageIncompleteDetailsReasonContentFilter MessageIncompleteDetailsReason = "content_filter"`

2427

2428 - `const MessageIncompleteDetailsReasonMaxTokens MessageIncompleteDetailsReason = "max_tokens"`

2429

2430 - `const MessageIncompleteDetailsReasonRunCancelled MessageIncompleteDetailsReason = "run_cancelled"`

2431

2432 - `const MessageIncompleteDetailsReasonRunExpired MessageIncompleteDetailsReason = "run_expired"`

2433

2434 - `const MessageIncompleteDetailsReasonRunFailed MessageIncompleteDetailsReason = "run_failed"`

2435

2436 - `Metadata Metadata`

2437

2438 Set of 16 key-value pairs that can be attached to an object. This can be

2439 useful for storing additional information about the object in a structured

2440 format, and querying for objects via API or the dashboard.

2441

2442 Keys are strings with a maximum length of 64 characters. Values are strings

2443 with a maximum length of 512 characters.

2444

2445 - `Object ThreadMessage`

2446

2447 The object type, which is always `thread.message`.

2448

2449 - `const ThreadMessageThreadMessage ThreadMessage = "thread.message"`

2450

2451 - `Role MessageRole`

2452

2453 The entity that produced the message. One of `user` or `assistant`.

2454

2455 - `const MessageRoleUser MessageRole = "user"`

2456

2457 - `const MessageRoleAssistant MessageRole = "assistant"`

2458

2459 - `RunID string`

2460

2461 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.

2462

2463 - `Status MessageStatus`

2464

2465 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

2466

2467 - `const MessageStatusInProgress MessageStatus = "in_progress"`

2468

2469 - `const MessageStatusIncomplete MessageStatus = "incomplete"`

2470

2471 - `const MessageStatusCompleted MessageStatus = "completed"`

2472

2473 - `ThreadID string`

2474

2475 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.

2476

2477 - `Event ThreadMessageCreated`

2478

2479 - `const ThreadMessageCreatedThreadMessageCreated ThreadMessageCreated = "thread.message.created"`

2480

2481 - `type AssistantStreamEventThreadMessageInProgress struct{…}`

2482

2483 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) moves to an `in_progress` state.

2484

2485 - `Data Message`

2486

2487 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2488

2489 - `Event ThreadMessageInProgress`

2490

2491 - `const ThreadMessageInProgressThreadMessageInProgress ThreadMessageInProgress = "thread.message.in_progress"`

2492

2493 - `type AssistantStreamEventThreadMessageDelta struct{…}`

2494

2495 Occurs when parts of a [Message](https://platform.openai.com/docs/api-reference/messages/object) are being streamed.

2496

2497 - `Data MessageDeltaEvent`

2498

2499 Represents a message delta i.e. any changed fields on a message during streaming.

2500

2501 - `ID string`

2502

2503 The identifier of the message, which can be referenced in API endpoints.

2504

2505 - `Delta MessageDelta`

2506

2507 The delta containing the fields that have changed on the Message.

2508

2509 - `Content []any`

2510

2511 The content of the message in array of text and/or images.

2512

2513 - `Role MessageDeltaRole`

2514

2515 The entity that produced the message. One of `user` or `assistant`.

2516

2517 - `const MessageDeltaRoleUser MessageDeltaRole = "user"`

2518

2519 - `const MessageDeltaRoleAssistant MessageDeltaRole = "assistant"`

2520

2521 - `Object ThreadMessageDelta`

2522

2523 The object type, which is always `thread.message.delta`.

2524

2525 - `const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"`

2526

2527 - `Event ThreadMessageDelta`

2528

2529 - `const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"`

2530

2531 - `type AssistantStreamEventThreadMessageCompleted struct{…}`

2532

2533 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is completed.

2534

2535 - `Data Message`

2536

2537 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2538

2539 - `Event ThreadMessageCompleted`

2540

2541 - `const ThreadMessageCompletedThreadMessageCompleted ThreadMessageCompleted = "thread.message.completed"`

2542

2543 - `type AssistantStreamEventThreadMessageIncomplete struct{…}`

2544

2545 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) ends before it is completed.

2546

2547 - `Data Message`

2548

2549 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2550

2551 - `Event ThreadMessageIncomplete`

2552

2553 - `const ThreadMessageIncompleteThreadMessageIncomplete ThreadMessageIncomplete = "thread.message.incomplete"`

2554

2555 - `type AssistantStreamEventErrorEvent struct{…}`

2556

2557 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.

2558

2559 - `Data ErrorObject`

2560

2561 - `Code string`

2562

2563 - `Message string`

2564

2565 - `Param string`

2566

2567 - `Type string`

2568

2569 - `Event Error`

2570

2571 - `const ErrorError Error = "error"`

2572

2573### Assistant Tool

2574

2575- `type AssistantTool interface{…}`

2576

2577### Code Interpreter Tool

2578

2579- `type CodeInterpreterTool struct{…}`

2580

2581 - `Type CodeInterpreter`

2582

2583 The type of tool being defined: `code_interpreter`

2584

2585 - `const CodeInterpreterCodeInterpreter CodeInterpreter = "code_interpreter"`

2586

2587### File Search Tool

2588

2589- `type FileSearchTool struct{…}`

2590

2591 - `Type FileSearch`

2592

2593 The type of tool being defined: `file_search`

2594

2595 - `const FileSearchFileSearch FileSearch = "file_search"`

2596

2597 - `FileSearch FileSearchToolFileSearch`

2598

2599 Overrides for the file search tool.

2600

2601 - `MaxNumResults int64`

2602

2603 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.

2604

2605 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.

2606

2607 - `RankingOptions FileSearchToolFileSearchRankingOptions`

2608

2609 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.

2610

2611 See the [file search tool documentation](https://platform.openai.com/docs/assistants/tools/file-search#customizing-file-search-settings) for more information.

2612

2613 - `ScoreThreshold float64`

2614

2615 The score threshold for the file search. All values must be a floating point number between 0 and 1.

2616

2617 - `Ranker string`

2618

2619 The ranker to use for the file search. If not specified will use the `auto` ranker.

2620

2621 - `const FileSearchToolFileSearchRankingOptionsRankerAuto FileSearchToolFileSearchRankingOptionsRanker = "auto"`

2622

2623 - `const FileSearchToolFileSearchRankingOptionsRankerDefault2024_08_21 FileSearchToolFileSearchRankingOptionsRanker = "default_2024_08_21"`

2624

2625### Function Tool

2626

2627- `type FunctionTool struct{…}`

2628

2629 - `Function FunctionDefinition`

2630

2631 - `Name string`

2632

2633 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.

2634

2635 - `Description string`

2636

2637 A description of what the function does, used by the model to choose when and how to call the function.

2638

2639 - `Parameters FunctionParameters`

2640

2641 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.

2642

2643 Omitting `parameters` defines a function with an empty parameter list.

2644

2645 - `Strict bool`

2646

2647 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).

2648

2649 - `Type Function`

2650

2651 The type of tool being defined: `function`

2652

2653 - `const FunctionFunction Function = "function"`

2654

2655### Message Stream Event

2656

2657- `type MessageStreamEventUnion interface{…}`

2658

2659 Occurs when a [message](https://platform.openai.com/docs/api-reference/messages/object) is created.

2660

2661 - `MessageStreamEventThreadMessageCreated`

2662

2663 - `Data Message`

2664

2665 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2666

2667 - `ID string`

2668

2669 The identifier, which can be referenced in API endpoints.

2670

2671 - `AssistantID string`

2672

2673 If applicable, the ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) that authored this message.

2674

2675 - `Attachments []MessageAttachment`

2676

2677 A list of files attached to the message, and the tools they were added to.

2678

2679 - `FileID string`

2680

2681 The ID of the file to attach to the message.

2682

2683 - `Tools []MessageAttachmentToolUnion`

2684

2685 The tools to add this file to.

2686

2687 - `type CodeInterpreterTool struct{…}`

2688

2689 - `Type CodeInterpreter`

2690

2691 The type of tool being defined: `code_interpreter`

2692

2693 - `const CodeInterpreterCodeInterpreter CodeInterpreter = "code_interpreter"`

2694

2695 - `type MessageAttachmentToolAssistantToolsFileSearchTypeOnly struct{…}`

2696

2697 - `Type FileSearch`

2698

2699 The type of tool being defined: `file_search`

2700

2701 - `const FileSearchFileSearch FileSearch = "file_search"`

2702

2703 - `CompletedAt int64`

2704

2705 The Unix timestamp (in seconds) for when the message was completed.

2706

2707 - `Content []any`

2708

2709 The content of the message in array of text and/or images.

2710

2711 - `CreatedAt int64`

2712

2713 The Unix timestamp (in seconds) for when the message was created.

2714

2715 - `IncompleteAt int64`

2716

2717 The Unix timestamp (in seconds) for when the message was marked as incomplete.

2718

2719 - `IncompleteDetails MessageIncompleteDetails`

2720

2721 On an incomplete message, details about why the message is incomplete.

2722

2723 - `Reason string`

2724

2725 The reason the message is incomplete.

2726

2727 - `const MessageIncompleteDetailsReasonContentFilter MessageIncompleteDetailsReason = "content_filter"`

2728

2729 - `const MessageIncompleteDetailsReasonMaxTokens MessageIncompleteDetailsReason = "max_tokens"`

2730

2731 - `const MessageIncompleteDetailsReasonRunCancelled MessageIncompleteDetailsReason = "run_cancelled"`

2732

2733 - `const MessageIncompleteDetailsReasonRunExpired MessageIncompleteDetailsReason = "run_expired"`

2734

2735 - `const MessageIncompleteDetailsReasonRunFailed MessageIncompleteDetailsReason = "run_failed"`

2736

2737 - `Metadata Metadata`

2738

2739 Set of 16 key-value pairs that can be attached to an object. This can be

2740 useful for storing additional information about the object in a structured

2741 format, and querying for objects via API or the dashboard.

2742

2743 Keys are strings with a maximum length of 64 characters. Values are strings

2744 with a maximum length of 512 characters.

2745

2746 - `Object ThreadMessage`

2747

2748 The object type, which is always `thread.message`.

2749

2750 - `const ThreadMessageThreadMessage ThreadMessage = "thread.message"`

2751

2752 - `Role MessageRole`

2753

2754 The entity that produced the message. One of `user` or `assistant`.

2755

2756 - `const MessageRoleUser MessageRole = "user"`

2757

2758 - `const MessageRoleAssistant MessageRole = "assistant"`

2759

2760 - `RunID string`

2761

2762 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.

2763

2764 - `Status MessageStatus`

2765

2766 The status of the message, which can be either `in_progress`, `incomplete`, or `completed`.

2767

2768 - `const MessageStatusInProgress MessageStatus = "in_progress"`

2769

2770 - `const MessageStatusIncomplete MessageStatus = "incomplete"`

2771

2772 - `const MessageStatusCompleted MessageStatus = "completed"`

2773

2774 - `ThreadID string`

2775

2776 The [thread](https://platform.openai.com/docs/api-reference/threads) ID that this message belongs to.

2777

2778 - `Event ThreadMessageCreated`

2779

2780 - `const ThreadMessageCreatedThreadMessageCreated ThreadMessageCreated = "thread.message.created"`

2781

2782 - `MessageStreamEventThreadMessageInProgress`

2783

2784 - `Data Message`

2785

2786 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2787

2788 - `Event ThreadMessageInProgress`

2789

2790 - `const ThreadMessageInProgressThreadMessageInProgress ThreadMessageInProgress = "thread.message.in_progress"`

2791

2792 - `MessageStreamEventThreadMessageDelta`

2793

2794 - `Data MessageDeltaEvent`

2795

2796 Represents a message delta i.e. any changed fields on a message during streaming.

2797

2798 - `ID string`

2799

2800 The identifier of the message, which can be referenced in API endpoints.

2801

2802 - `Delta MessageDelta`

2803

2804 The delta containing the fields that have changed on the Message.

2805

2806 - `Content []any`

2807

2808 The content of the message in array of text and/or images.

2809

2810 - `Role MessageDeltaRole`

2811

2812 The entity that produced the message. One of `user` or `assistant`.

2813

2814 - `const MessageDeltaRoleUser MessageDeltaRole = "user"`

2815

2816 - `const MessageDeltaRoleAssistant MessageDeltaRole = "assistant"`

2817

2818 - `Object ThreadMessageDelta`

2819

2820 The object type, which is always `thread.message.delta`.

2821

2822 - `const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"`

2823

2824 - `Event ThreadMessageDelta`

2825

2826 - `const ThreadMessageDeltaThreadMessageDelta ThreadMessageDelta = "thread.message.delta"`

2827

2828 - `MessageStreamEventThreadMessageCompleted`

2829

2830 - `Data Message`

2831

2832 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2833

2834 - `Event ThreadMessageCompleted`

2835

2836 - `const ThreadMessageCompletedThreadMessageCompleted ThreadMessageCompleted = "thread.message.completed"`

2837

2838 - `MessageStreamEventThreadMessageIncomplete`

2839

2840 - `Data Message`

2841

2842 Represents a message within a [thread](https://platform.openai.com/docs/api-reference/threads).

2843

2844 - `Event ThreadMessageIncomplete`

2845

2846 - `const ThreadMessageIncompleteThreadMessageIncomplete ThreadMessageIncomplete = "thread.message.incomplete"`

2847

2848### Run Step Stream Event

2849

2850- `type RunStepStreamEventUnion interface{…}`

2851

2852 Occurs when a [run step](https://platform.openai.com/docs/api-reference/run-steps/step-object) is created.

2853

2854 - `RunStepStreamEventThreadRunStepCreated`

2855

2856 - `Data RunStep`

2857

2858 Represents a step in execution of a run.

2859

2860 - `ID string`

2861

2862 The identifier of the run step, which can be referenced in API endpoints.

2863

2864 - `AssistantID string`

2865

2866 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) associated with the run step.

2867

2868 - `CancelledAt int64`

2869

2870 The Unix timestamp (in seconds) for when the run step was cancelled.

2871

2872 - `CompletedAt int64`

2873

2874 The Unix timestamp (in seconds) for when the run step completed.

2875

2876 - `CreatedAt int64`

2877

2878 The Unix timestamp (in seconds) for when the run step was created.

2879

2880 - `ExpiredAt int64`

2881

2882 The Unix timestamp (in seconds) for when the run step expired. A step is considered expired if the parent run is expired.

2883

2884 - `FailedAt int64`

2885

2886 The Unix timestamp (in seconds) for when the run step failed.

2887

2888 - `LastError RunStepLastError`

2889

2890 The last error associated with this run step. Will be `null` if there are no errors.

2891

2892 - `Code string`

2893

2894 One of `server_error` or `rate_limit_exceeded`.

2895

2896 - `const RunStepLastErrorCodeServerError RunStepLastErrorCode = "server_error"`

2897

2898 - `const RunStepLastErrorCodeRateLimitExceeded RunStepLastErrorCode = "rate_limit_exceeded"`

2899

2900 - `Message string`

2901

2902 A human-readable description of the error.

2903

2904 - `Metadata Metadata`

2905

2906 Set of 16 key-value pairs that can be attached to an object. This can be

2907 useful for storing additional information about the object in a structured

2908 format, and querying for objects via API or the dashboard.

2909

2910 Keys are strings with a maximum length of 64 characters. Values are strings

2911 with a maximum length of 512 characters.

2912

2913 - `Object ThreadRunStep`

2914

2915 The object type, which is always `thread.run.step`.

2916

2917 - `const ThreadRunStepThreadRunStep ThreadRunStep = "thread.run.step"`

2918

2919 - `RunID string`

2920

2921 The ID of the [run](https://platform.openai.com/docs/api-reference/runs) that this run step is a part of.

2922

2923 - `Status RunStepStatus`

2924

2925 The status of the run step, which can be either `in_progress`, `cancelled`, `failed`, `completed`, or `expired`.

2926

2927 - `const RunStepStatusInProgress RunStepStatus = "in_progress"`

2928

2929 - `const RunStepStatusCancelled RunStepStatus = "cancelled"`

2930

2931 - `const RunStepStatusFailed RunStepStatus = "failed"`

2932

2933 - `const RunStepStatusCompleted RunStepStatus = "completed"`

2934

2935 - `const RunStepStatusExpired RunStepStatus = "expired"`

2936

2937 - `StepDetails RunStepStepDetailsUnion`

2938

2939 The details of the run step.

2940

2941 - `type MessageCreationStepDetails struct{…}`

2942

2943 Details of the message creation by the run step.

2944

2945 - `MessageCreation MessageCreationStepDetailsMessageCreation`

2946

2947 - `MessageID string`

2948

2949 The ID of the message that was created by this run step.

2950

2951 - `Type MessageCreation`

2952

2953 Always `message_creation`.

2954

2955 - `const MessageCreationMessageCreation MessageCreation = "message_creation"`

2956

2957 - `type ToolCallsStepDetails struct{…}`

2958

2959 Details of the tool call.

2960

2961 - `ToolCalls []any`

2962

2963 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`.

2964

2965 - `Type ToolCalls`

2966

2967 Always `tool_calls`.

2968

2969 - `const ToolCallsToolCalls ToolCalls = "tool_calls"`

2970

2971 - `ThreadID string`

2972

2973 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was run.

2974

2975 - `Type RunStepType`

2976

2977 The type of run step, which can be either `message_creation` or `tool_calls`.

2978

2979 - `const RunStepTypeMessageCreation RunStepType = "message_creation"`

2980

2981 - `const RunStepTypeToolCalls RunStepType = "tool_calls"`

2982

2983 - `Usage RunStepUsage`

2984

2985 Usage statistics related to the run step. This value will be `null` while the run step's status is `in_progress`.

2986

2987 - `CompletionTokens int64`

2988

2989 Number of completion tokens used over the course of the run step.

2990

2991 - `PromptTokens int64`

2992

2993 Number of prompt tokens used over the course of the run step.

2994

2995 - `TotalTokens int64`

2996

2997 Total number of tokens used (prompt + completion).

2998

2999 - `Event ThreadRunStepCreated`

3000

3001 - `const ThreadRunStepCreatedThreadRunStepCreated ThreadRunStepCreated = "thread.run.step.created"`

3002

3003 - `RunStepStreamEventThreadRunStepInProgress`

3004

3005 - `Data RunStep`

3006

3007 Represents a step in execution of a run.

3008

3009 - `Event ThreadRunStepInProgress`

3010

3011 - `const ThreadRunStepInProgressThreadRunStepInProgress ThreadRunStepInProgress = "thread.run.step.in_progress"`

3012

3013 - `RunStepStreamEventThreadRunStepDelta`

3014

3015 - `Data RunStepDeltaEvent`

3016

3017 Represents a run step delta i.e. any changed fields on a run step during streaming.

3018

3019 - `ID string`

3020

3021 The identifier of the run step, which can be referenced in API endpoints.

3022

3023 - `Delta any`

3024

3025 - `Object ThreadRunStepDelta`

3026

3027 The object type, which is always `thread.run.step.delta`.

3028

3029 - `const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"`

3030

3031 - `Event ThreadRunStepDelta`

3032

3033 - `const ThreadRunStepDeltaThreadRunStepDelta ThreadRunStepDelta = "thread.run.step.delta"`

3034

3035 - `RunStepStreamEventThreadRunStepCompleted`

3036

3037 - `Data RunStep`

3038

3039 Represents a step in execution of a run.

3040

3041 - `Event ThreadRunStepCompleted`

3042

3043 - `const ThreadRunStepCompletedThreadRunStepCompleted ThreadRunStepCompleted = "thread.run.step.completed"`

3044

3045 - `RunStepStreamEventThreadRunStepFailed`

3046

3047 - `Data RunStep`

3048

3049 Represents a step in execution of a run.

3050

3051 - `Event ThreadRunStepFailed`

3052

3053 - `const ThreadRunStepFailedThreadRunStepFailed ThreadRunStepFailed = "thread.run.step.failed"`

3054

3055 - `RunStepStreamEventThreadRunStepCancelled`

3056

3057 - `Data RunStep`

3058

3059 Represents a step in execution of a run.

3060

3061 - `Event ThreadRunStepCancelled`

3062

3063 - `const ThreadRunStepCancelledThreadRunStepCancelled ThreadRunStepCancelled = "thread.run.step.cancelled"`

3064

3065 - `RunStepStreamEventThreadRunStepExpired`

3066

3067 - `Data RunStep`

3068

3069 Represents a step in execution of a run.

3070

3071 - `Event ThreadRunStepExpired`

3072

3073 - `const ThreadRunStepExpiredThreadRunStepExpired ThreadRunStepExpired = "thread.run.step.expired"`

3074

3075### Run Stream Event

3076

3077- `type RunStreamEventUnion interface{…}`

3078

3079 Occurs when a new [run](https://platform.openai.com/docs/api-reference/runs/object) is created.

3080

3081 - `RunStreamEventThreadRunCreated`

3082

3083 - `Data Run`

3084

3085 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3086

3087 - `ID string`

3088

3089 The identifier, which can be referenced in API endpoints.

3090

3091 - `AssistantID string`

3092

3093 The ID of the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for execution of this run.

3094

3095 - `CancelledAt int64`

3096

3097 The Unix timestamp (in seconds) for when the run was cancelled.

3098

3099 - `CompletedAt int64`

3100

3101 The Unix timestamp (in seconds) for when the run was completed.

3102

3103 - `CreatedAt int64`

3104

3105 The Unix timestamp (in seconds) for when the run was created.

3106

3107 - `ExpiresAt int64`

3108

3109 The Unix timestamp (in seconds) for when the run will expire.

3110

3111 - `FailedAt int64`

3112

3113 The Unix timestamp (in seconds) for when the run failed.

3114

3115 - `IncompleteDetails RunIncompleteDetails`

3116

3117 Details on why the run is incomplete. Will be `null` if the run is not incomplete.

3118

3119 - `Reason string`

3120

3121 The reason why the run is incomplete. This will point to which specific token limit was reached over the course of the run.

3122

3123 - `const RunIncompleteDetailsReasonMaxCompletionTokens RunIncompleteDetailsReason = "max_completion_tokens"`

3124

3125 - `const RunIncompleteDetailsReasonMaxPromptTokens RunIncompleteDetailsReason = "max_prompt_tokens"`

3126

3127 - `Instructions string`

3128

3129 The instructions that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

3130

3131 - `LastError RunLastError`

3132

3133 The last error associated with this run. Will be `null` if there are no errors.

3134

3135 - `Code string`

3136

3137 One of `server_error`, `rate_limit_exceeded`, or `invalid_prompt`.

3138

3139 - `const RunLastErrorCodeServerError RunLastErrorCode = "server_error"`

3140

3141 - `const RunLastErrorCodeRateLimitExceeded RunLastErrorCode = "rate_limit_exceeded"`

3142

3143 - `const RunLastErrorCodeInvalidPrompt RunLastErrorCode = "invalid_prompt"`

3144

3145 - `Message string`

3146

3147 A human-readable description of the error.

3148

3149 - `MaxCompletionTokens int64`

3150

3151 The maximum number of completion tokens specified to have been used over the course of the run.

3152

3153 - `MaxPromptTokens int64`

3154

3155 The maximum number of prompt tokens specified to have been used over the course of the run.

3156

3157 - `Metadata Metadata`

3158

3159 Set of 16 key-value pairs that can be attached to an object. This can be

3160 useful for storing additional information about the object in a structured

3161 format, and querying for objects via API or the dashboard.

3162

3163 Keys are strings with a maximum length of 64 characters. Values are strings

3164 with a maximum length of 512 characters.

3165

3166 - `Model string`

3167

3168 The model that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

3169

3170 - `Object ThreadRun`

3171

3172 The object type, which is always `thread.run`.

3173

3174 - `const ThreadRunThreadRun ThreadRun = "thread.run"`

3175

3176 - `ParallelToolCalls bool`

3177

3178 Whether to enable [parallel function calling](https://platform.openai.com/docs/guides/function-calling#configuring-parallel-function-calling) during tool use.

3179

3180 - `RequiredAction RunRequiredAction`

3181

3182 Details on the action required to continue the run. Will be `null` if no action is required.

3183

3184 - `SubmitToolOutputs RunRequiredActionSubmitToolOutputs`

3185

3186 Details on the tool outputs needed for this run to continue.

3187

3188 - `ToolCalls []RequiredActionFunctionToolCall`

3189

3190 A list of the relevant tool calls.

3191

3192 - `ID string`

3193

3194 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](https://platform.openai.com/docs/api-reference/runs/submitToolOutputs) endpoint.

3195

3196 - `Function RequiredActionFunctionToolCallFunction`

3197

3198 The function definition.

3199

3200 - `Arguments string`

3201

3202 The arguments that the model expects you to pass to the function.

3203

3204 - `Name string`

3205

3206 The name of the function.

3207

3208 - `Type Function`

3209

3210 The type of tool call the output is required for. For now, this is always `function`.

3211

3212 - `const FunctionFunction Function = "function"`

3213

3214 - `Type SubmitToolOutputs`

3215

3216 For now, this is always `submit_tool_outputs`.

3217

3218 - `const SubmitToolOutputsSubmitToolOutputs SubmitToolOutputs = "submit_tool_outputs"`

3219

3220 - `ResponseFormat AssistantResponseFormatOptionUnion`

3221

3222 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`.

3223

3224 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).

3225

3226 Setting to `{ "type": "json_object" }` enables JSON mode, which ensures the message the model generates is valid JSON.

3227

3228 **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.

3229

3230 - `type Auto string`

3231

3232 `auto` is the default value

3233

3234 - `const AutoAuto Auto = "auto"`

3235

3236 - `type ResponseFormatText struct{…}`

3237

3238 Default response format. Used to generate text responses.

3239

3240 - `Type Text`

3241

3242 The type of response format being defined. Always `text`.

3243

3244 - `const TextText Text = "text"`

3245

3246 - `type ResponseFormatJSONObject struct{…}`

3247

3248 JSON object response format. An older method of generating JSON responses.

3249 Using `json_schema` is recommended for models that support it. Note that the

3250 model will not generate JSON without a system or user message instructing it

3251 to do so.

3252

3253 - `Type JSONObject`

3254

3255 The type of response format being defined. Always `json_object`.

3256

3257 - `const JSONObjectJSONObject JSONObject = "json_object"`

3258

3259 - `type ResponseFormatJSONSchema struct{…}`

3260

3261 JSON Schema response format. Used to generate structured JSON responses.

3262 Learn more about [Structured Outputs](https://platform.openai.com/docs/guides/structured-outputs).

3263

3264 - `JSONSchema ResponseFormatJSONSchemaJSONSchema`

3265

3266 Structured Outputs configuration options, including a JSON Schema.

3267

3268 - `Name string`

3269

3270 The name of the response format. Must be a-z, A-Z, 0-9, or contain

3271 underscores and dashes, with a maximum length of 64.

3272

3273 - `Description string`

3274

3275 A description of what the response format is for, used by the model to

3276 determine how to respond in the format.

3277

3278 - `Schema map[string, any]`

3279

3280 The schema for the response format, described as a JSON Schema object.

3281 Learn how to build JSON schemas [here](https://json-schema.org/).

3282

3283 - `Strict bool`

3284

3285 Whether to enable strict schema adherence when generating the output.

3286 If set to true, the model will always follow the exact schema defined

3287 in the `schema` field. Only a subset of JSON Schema is supported when

3288 `strict` is `true`. To learn more, read the [Structured Outputs

3289 guide](https://platform.openai.com/docs/guides/structured-outputs).

3290

3291 - `Type JSONSchema`

3292

3293 The type of response format being defined. Always `json_schema`.

3294

3295 - `const JSONSchemaJSONSchema JSONSchema = "json_schema"`

3296

3297 - `StartedAt int64`

3298

3299 The Unix timestamp (in seconds) for when the run was started.

3300

3301 - `Status any`

3302

3303 - `ThreadID string`

3304

3305 The ID of the [thread](https://platform.openai.com/docs/api-reference/threads) that was executed on as a part of this run.

3306

3307 - `ToolChoice AssistantToolChoiceOptionUnion`

3308

3309 Controls which (if any) tool is called by the model.

3310 `none` means the model will not call any tools and instead generates a message.

3311 `auto` is the default value and means the model can pick between generating a message or calling one or more tools.

3312 `required` means the model must call one or more tools before responding to the user.

3313 Specifying a particular tool like `{"type": "file_search"}` or `{"type": "function", "function": {"name": "my_function"}}` forces the model to call that tool.

3314

3315 - `type AssistantToolChoiceOptionAuto string`

3316

3317 `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.

3318

3319 - `const AssistantToolChoiceOptionAutoNone AssistantToolChoiceOptionAuto = "none"`

3320

3321 - `const AssistantToolChoiceOptionAutoAuto AssistantToolChoiceOptionAuto = "auto"`

3322

3323 - `const AssistantToolChoiceOptionAutoRequired AssistantToolChoiceOptionAuto = "required"`

3324

3325 - `type AssistantToolChoice struct{…}`

3326

3327 Specifies a tool the model should use. Use to force the model to call a specific tool.

3328

3329 - `Type AssistantToolChoiceType`

3330

3331 The type of the tool. If type is `function`, the function name must be set

3332

3333 - `const AssistantToolChoiceTypeFunction AssistantToolChoiceType = "function"`

3334

3335 - `const AssistantToolChoiceTypeCodeInterpreter AssistantToolChoiceType = "code_interpreter"`

3336

3337 - `const AssistantToolChoiceTypeFileSearch AssistantToolChoiceType = "file_search"`

3338

3339 - `Function AssistantToolChoiceFunction`

3340

3341 - `Name string`

3342

3343 The name of the function to call.

3344

3345 - `Tools []any`

3346

3347 The list of tools that the [assistant](https://platform.openai.com/docs/api-reference/assistants) used for this run.

3348

3349 - `TruncationStrategy RunTruncationStrategy`

3350

3351 Controls for how a thread will be truncated prior to the run. Use this to control the initial context window of the run.

3352

3353 - `Type string`

3354

3355 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`.

3356

3357 - `const RunTruncationStrategyTypeAuto RunTruncationStrategyType = "auto"`

3358

3359 - `const RunTruncationStrategyTypeLastMessages RunTruncationStrategyType = "last_messages"`

3360

3361 - `LastMessages int64`

3362

3363 The number of most recent messages from the thread when constructing the context for the run.

3364

3365 - `Usage RunUsage`

3366

3367 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.).

3368

3369 - `CompletionTokens int64`

3370

3371 Number of completion tokens used over the course of the run.

3372

3373 - `PromptTokens int64`

3374

3375 Number of prompt tokens used over the course of the run.

3376

3377 - `TotalTokens int64`

3378

3379 Total number of tokens used (prompt + completion).

3380

3381 - `Temperature float64`

3382

3383 The sampling temperature used for this run. If not set, defaults to 1.

3384

3385 - `TopP float64`

3386

3387 The nucleus sampling value used for this run. If not set, defaults to 1.

3388

3389 - `Event ThreadRunCreated`

3390

3391 - `const ThreadRunCreatedThreadRunCreated ThreadRunCreated = "thread.run.created"`

3392

3393 - `RunStreamEventThreadRunQueued`

3394

3395 - `Data Run`

3396

3397 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3398

3399 - `Event ThreadRunQueued`

3400

3401 - `const ThreadRunQueuedThreadRunQueued ThreadRunQueued = "thread.run.queued"`

3402

3403 - `RunStreamEventThreadRunInProgress`

3404

3405 - `Data Run`

3406

3407 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3408

3409 - `Event ThreadRunInProgress`

3410

3411 - `const ThreadRunInProgressThreadRunInProgress ThreadRunInProgress = "thread.run.in_progress"`

3412

3413 - `RunStreamEventThreadRunRequiresAction`

3414

3415 - `Data Run`

3416

3417 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3418

3419 - `Event ThreadRunRequiresAction`

3420

3421 - `const ThreadRunRequiresActionThreadRunRequiresAction ThreadRunRequiresAction = "thread.run.requires_action"`

3422

3423 - `RunStreamEventThreadRunCompleted`

3424

3425 - `Data Run`

3426

3427 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3428

3429 - `Event ThreadRunCompleted`

3430

3431 - `const ThreadRunCompletedThreadRunCompleted ThreadRunCompleted = "thread.run.completed"`

3432

3433 - `RunStreamEventThreadRunIncomplete`

3434

3435 - `Data Run`

3436

3437 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3438

3439 - `Event ThreadRunIncomplete`

3440

3441 - `const ThreadRunIncompleteThreadRunIncomplete ThreadRunIncomplete = "thread.run.incomplete"`

3442

3443 - `RunStreamEventThreadRunFailed`

3444

3445 - `Data Run`

3446

3447 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3448

3449 - `Event ThreadRunFailed`

3450

3451 - `const ThreadRunFailedThreadRunFailed ThreadRunFailed = "thread.run.failed"`

3452

3453 - `RunStreamEventThreadRunCancelling`

3454

3455 - `Data Run`

3456

3457 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3458

3459 - `Event ThreadRunCancelling`

3460

3461 - `const ThreadRunCancellingThreadRunCancelling ThreadRunCancelling = "thread.run.cancelling"`

3462

3463 - `RunStreamEventThreadRunCancelled`

3464

3465 - `Data Run`

3466

3467 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3468

3469 - `Event ThreadRunCancelled`

3470

3471 - `const ThreadRunCancelledThreadRunCancelled ThreadRunCancelled = "thread.run.cancelled"`

3472

3473 - `RunStreamEventThreadRunExpired`

3474

3475 - `Data Run`

3476

3477 Represents an execution run on a [thread](https://platform.openai.com/docs/api-reference/threads).

3478

3479 - `Event ThreadRunExpired`

3480

3481 - `const ThreadRunExpiredThreadRunExpired ThreadRunExpired = "thread.run.expired"`

3482

3483### Thread Stream Event

3484

3485- `type ThreadStreamEvent struct{…}`

3486

3487 Occurs when a new [thread](https://platform.openai.com/docs/api-reference/threads/object) is created.

3488

3489 - `Data Thread`

3490

3491 Represents a thread that contains [messages](https://platform.openai.com/docs/api-reference/messages).

3492

3493 - `ID string`

3494

3495 The identifier, which can be referenced in API endpoints.

3496

3497 - `CreatedAt int64`

3498

3499 The Unix timestamp (in seconds) for when the thread was created.

3500

3501 - `Metadata Metadata`

3502

3503 Set of 16 key-value pairs that can be attached to an object. This can be

3504 useful for storing additional information about the object in a structured

3505 format, and querying for objects via API or the dashboard.

3506

3507 Keys are strings with a maximum length of 64 characters. Values are strings

3508 with a maximum length of 512 characters.

3509

3510 - `Object Thread`

3511

3512 The object type, which is always `thread`.

3513

3514 - `const ThreadThread Thread = "thread"`

3515

3516 - `ToolResources ThreadToolResources`

3517

3518 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.

3519

3520 - `CodeInterpreter ThreadToolResourcesCodeInterpreter`

3521

3522 - `FileIDs []string`

3523

3524 A list of [file](https://platform.openai.com/docs/api-reference/files) IDs made available to the `code_interpreter` tool. There can be a maximum of 20 files associated with the tool.

3525

3526 - `FileSearch ThreadToolResourcesFileSearch`

3527

3528 - `VectorStoreIDs []string`

3529

3530 The [vector store](https://platform.openai.com/docs/api-reference/vector-stores/object) attached to this thread. There can be a maximum of 1 vector store attached to the thread.

3531

3532 - `Event ThreadCreated`

3533

3534 - `const ThreadCreatedThreadCreated ThreadCreated = "thread.created"`

3535

3536 - `Enabled bool`

3537

3538 Whether to enable input audio transcription.

go/resources/beta/subresources/assistants/index.md 2026-07-07 08:02 UTC to 2026-07-09 20:58 UTC

Assistants

List assistants

Parameters

Returns

Example

Response

Create assistant

Parameters

Returns

Example

Response

Retrieve assistant

Parameters

Returns

Example

Response

Modify assistant

Parameters

Returns

Example

Response

Delete assistant

Parameters

Returns

Example

Response

Domain Types

Assistant

Assistant Deleted

Assistant Stream Event

Assistant Tool

Code Interpreter Tool

File Search Tool

Function Tool

Message Stream Event

Run Step Stream Event

Run Stream Event

Thread Stream Event