Creates a model response. Provide text or image inputs to generate text or JSON outputs. Have the model call your own custom code or use built-in tools like web search or file search to use your own data as input for the model’s response.

Opaque per-user identity string used to scope endpoint-scoped data to a specific end user. The caller must have the agents/endpoints/UserIdentityImpersonation/action RBAC permission.

An opaque, application-owned conversation identifier. The value is not returned in the response and must not contain secrets or personally identifiable information.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Domain-specific access programs to use for this request.

Options for prompt caching. Supported for gpt-5.6 and later models. By default, OpenAI automatically chooses one implicit cache breakpoint. You can add explicit breakpoints to content blocks with prompt_cache_breakpoint. Each request can write up to four breakpoints. For cache matching, OpenAI considers up to the latest 80 breakpoints in the conversation, without a content-block lookback limit. Set mode to explicit to disable the implicit breakpoint. The ttl defaults to 30m, which is currently the only supported value. See the prompt caching guide for current details.

Prepares the prompt cache without generating output. Defaults to false. When set to true, overrides the generate field to false.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Configuration options for reasoning models.

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.

Text, image, or file inputs to the model, used to generate a response. Learn more: - Text inputs and outputs - Image inputs - File inputs - Conversation state - Function calling

Configuration for running moderation on the input and output of this response.

The moderation model to use for moderated completions, e.g. 'omni-moderation-latest'.

The policy to apply to moderated response input and output.

The moderation policy for the response input.

The moderation policy for the response input.

Options for streaming responses. Only set this when you set stream: true.

When true, stream obfuscation will be enabled. Stream obfuscation adds random characters to an obfuscation field on streaming delta events to normalize payload sizes as a mitigation to certain side-channel attacks. These obfuscation fields are included by default, but add a small amount of overhead to the data stream. You can set include_obfuscation to false to optimize for bandwidth if you trust the network links between your application and the OpenAI API.

The conversation that this response belongs to. Items from this conversation are prepended to input_items for this response request. Input items and output items from this response are automatically added to this conversation after this response completes.

The unique ID of the conversation.

Context management configuration for this request.

Optional session identifier for sandbox affinity. Currently only relevant for hosted agents. When provided, the request is routed to the same sandbox. When omitted, auto-derived from conversation_id/prev_response_id or a new UUID is generated.

The name of the agent.

The version identifier of the agent.

The structured inputs to the response that can participate in prompt template substitution or tool argument bindings.

User security context contains several parameters that describe the application itself, and the end user that interacts with the application. These fields assist your security operations teams to investigate and mitigate security incidents by providing a comprehensive approach to protecting your AI applications. Learn more about protecting AI applications using Microsoft Defender for Cloud.

The name of the application. Sensitive personal information should not be included in this field.

This identifier is the Microsoft Entra ID (formerly Azure Active Directory) user object ID used to authenticate end-users within the generative AI application. Sensitive personal information should not be included in this field.

The Microsoft 365 tenant ID the end user belongs to. It's required when the generative AI application is multitenant.

Captures the original client's IP address.

Per-request configuration for Model Router inference calls. Supported on Model Router endpoints (deployed or instant if supported). Returns a 400 (Bad Request) error when the target is not a Model Router endpoint.

Request settings that control Model Router session affinity.

Request modes supported by Model Router session affinity.

An opaque, application-owned conversation identifier. Use the same value for related requests. The value is not returned in the response and must not contain secrets or personally identifiable information.

Session ID for this request. Only present for hosted agent responses. Returns the provided session ID or an auto-generated one if not provided in the request.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Unique identifier for this Response.

The object type of this resource - always set to response.

The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete.

Unix timestamp (in seconds) of when this Response was created.

An error object returned when the model fails to generate a Response.

The original error type, or null if it was not recorded.

The request parameter associated with the error, or null if none was recorded.

The error code for the response.

A human-readable description of the error.

The public explanation for this block.

The public continuation instruction.

An array of content items generated by the model. - The length and order of items in the output array is dependent on the model's response. - Rather than accessing the first item in the output array and assuming it's an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.

Configuration options for reasoning models.

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.

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

The prompt-caching options that were applied to the response. Supported for gpt-5.6 and later models.

Prompt cache diagnostics requested for this response.

Moderation results or errors for the response input and output.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Whether to allow the model to run tool calls in parallel.

The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation.

The unique ID of the conversation that this response was associated with.

The agent used for this response

The name of the agent.

The version identifier of the agent.

Model selection metadata providing visibility into routing decisions.

Routing decision metadata from Model Router.

The routing mode used by Model Router for the request.

Effective Model Router session affinity metadata for a request.

Effective modes reported for Model Router session affinity.

Conversation identifier sources supported by Model Router session affinity.

Final outcomes reported for Model Router session affinity.

Ordered trace of routing decisions. Contains one entry for Chat Completions, and may contain multiple entries for Responses API or agent scenarios where multiple inference turns occur.

The content filter evaluation results.

The type of the event. Always response.audio.delta.

A chunk of Base64 encoded response audio bytes.

The type of the event. Always response.audio.transcript.delta.

The partial transcript of the audio response.

The type of the event. Always response.code_interpreter_call_code.delta.

The unique identifier of the code interpreter tool call item.

The partial code snippet being streamed by the code interpreter.

The type of the event. Always response.code_interpreter_call.in_progress.

The unique identifier of the code interpreter tool call item.

The type of the event. Always response.code_interpreter_call.interpreting.

The unique identifier of the code interpreter tool call item.

The type of the event. Always response.content_part.added.

The ID of the output item that the content part was added to.

A text output from the model.

The type of the output text. Always output_text.

The text output from the model.

The annotations of the text output.

A refusal from the model.

The type of the refusal. Always refusal.

The refusal explanation from the model.

Reasoning text from the model.

The type of the reasoning text. Always reasoning_text.

The reasoning text from the model.

The type of the event. Always response.created.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Unique identifier for this Response.

The object type of this resource - always set to response.

The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete.

Unix timestamp (in seconds) of when this Response was created.

An error object returned when the model fails to generate a Response.

The original error type, or null if it was not recorded.

The request parameter associated with the error, or null if none was recorded.

The error code for the response.

A human-readable description of the error.

The public explanation for this block.

The public continuation instruction.

An array of content items generated by the model. - The length and order of items in the output array is dependent on the model's response. - Rather than accessing the first item in the output array and assuming it's an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.

Configuration options for reasoning models.

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.

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

The prompt-caching options that were applied to the response. Supported for gpt-5.6 and later models.

Prompt cache diagnostics requested for this response.

Moderation results or errors for the response input and output.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Whether to allow the model to run tool calls in parallel.

The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation.

The unique ID of the conversation that this response was associated with.

The agent used for this response

The name of the agent.

The version identifier of the agent.

Model selection metadata providing visibility into routing decisions.

Routing decision metadata from Model Router.

The routing mode used by Model Router for the request.

Effective Model Router session affinity metadata for a request.

Effective modes reported for Model Router session affinity.

Conversation identifier sources supported by Model Router session affinity.

Final outcomes reported for Model Router session affinity.

Ordered trace of routing decisions. Contains one entry for Chat Completions, and may contain multiple entries for Responses API or agent scenarios where multiple inference turns occur.

The content filter evaluation results.

The type of the event. Always error.

The error message.

The type of the event. Always response.file_search_call.in_progress.

The ID of the output item that the file search call is initiated.

The type of the event. Always response.file_search_call.searching.

The ID of the output item that the file search call is initiated.

The type of the event. Always response.function_call_arguments.delta.

The ID of the output item that the function-call arguments delta is added to.

The function-call arguments delta that is added.

The type of the event. Always response.in_progress.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Unique identifier for this Response.

The object type of this resource - always set to response.

The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete.

Unix timestamp (in seconds) of when this Response was created.

An error object returned when the model fails to generate a Response.

The original error type, or null if it was not recorded.

The request parameter associated with the error, or null if none was recorded.

The error code for the response.

A human-readable description of the error.

The public explanation for this block.

The public continuation instruction.

An array of content items generated by the model. - The length and order of items in the output array is dependent on the model's response. - Rather than accessing the first item in the output array and assuming it's an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.

Configuration options for reasoning models.

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.

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

The prompt-caching options that were applied to the response. Supported for gpt-5.6 and later models.

Prompt cache diagnostics requested for this response.

Moderation results or errors for the response input and output.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Whether to allow the model to run tool calls in parallel.

The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation.

The unique ID of the conversation that this response was associated with.

The agent used for this response

The name of the agent.

The version identifier of the agent.

Model selection metadata providing visibility into routing decisions.

Routing decision metadata from Model Router.

The routing mode used by Model Router for the request.

Effective Model Router session affinity metadata for a request.

Effective modes reported for Model Router session affinity.

Conversation identifier sources supported by Model Router session affinity.

Final outcomes reported for Model Router session affinity.

Ordered trace of routing decisions. Contains one entry for Chat Completions, and may contain multiple entries for Responses API or agent scenarios where multiple inference turns occur.

The content filter evaluation results.

The type of the event. Always response.failed.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Unique identifier for this Response.

The object type of this resource - always set to response.

The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete.

Unix timestamp (in seconds) of when this Response was created.

An error object returned when the model fails to generate a Response.

The original error type, or null if it was not recorded.

The request parameter associated with the error, or null if none was recorded.

The error code for the response.

A human-readable description of the error.

The public explanation for this block.

The public continuation instruction.

An array of content items generated by the model. - The length and order of items in the output array is dependent on the model's response. - Rather than accessing the first item in the output array and assuming it's an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.

Configuration options for reasoning models.

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.

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

The prompt-caching options that were applied to the response. Supported for gpt-5.6 and later models.

Prompt cache diagnostics requested for this response.

Moderation results or errors for the response input and output.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Whether to allow the model to run tool calls in parallel.

The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation.

The unique ID of the conversation that this response was associated with.

The agent used for this response

The name of the agent.

The version identifier of the agent.

Model selection metadata providing visibility into routing decisions.

Routing decision metadata from Model Router.

The routing mode used by Model Router for the request.

Effective Model Router session affinity metadata for a request.

Effective modes reported for Model Router session affinity.

Conversation identifier sources supported by Model Router session affinity.

Final outcomes reported for Model Router session affinity.

Ordered trace of routing decisions. Contains one entry for Chat Completions, and may contain multiple entries for Responses API or agent scenarios where multiple inference turns occur.

The content filter evaluation results.

The type of the event. Always response.incomplete.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Unique identifier for this Response.

The object type of this resource - always set to response.

The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete.

Unix timestamp (in seconds) of when this Response was created.

An error object returned when the model fails to generate a Response.

The original error type, or null if it was not recorded.

The request parameter associated with the error, or null if none was recorded.

The error code for the response.

A human-readable description of the error.

The public explanation for this block.

The public continuation instruction.

An array of content items generated by the model. - The length and order of items in the output array is dependent on the model's response. - Rather than accessing the first item in the output array and assuming it's an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.

Configuration options for reasoning models.

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.

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

The prompt-caching options that were applied to the response. Supported for gpt-5.6 and later models.

Prompt cache diagnostics requested for this response.

Moderation results or errors for the response input and output.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Whether to allow the model to run tool calls in parallel.

The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation.

The unique ID of the conversation that this response was associated with.

The agent used for this response

The name of the agent.

The version identifier of the agent.

Model selection metadata providing visibility into routing decisions.

Routing decision metadata from Model Router.

The routing mode used by Model Router for the request.

Effective Model Router session affinity metadata for a request.

Effective modes reported for Model Router session affinity.

Conversation identifier sources supported by Model Router session affinity.

Final outcomes reported for Model Router session affinity.

Ordered trace of routing decisions. Contains one entry for Chat Completions, and may contain multiple entries for Responses API or agent scenarios where multiple inference turns occur.

The content filter evaluation results.

The type of the event. Always response.output_item.added.

The structured output captured during the response.

The kind of CSDL action (e.g., 'SetVariable', 'InvokeAzureAgent').

Unique identifier for the action.

ID of the parent action if this is a nested action.

ID of the previous action if this action follows another.

Status of the action (e.g., 'in_progress', 'completed', 'failed', 'cancelled').

Request from the service for the user to perform OAuth consent.

The link the user can use to perform OAuth consent.

The server label for the OAuth consent request.

A Bing grounding tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a Bing grounding tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A SharePoint grounding tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a SharePoint grounding tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An Azure AI Search tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an Azure AI Search tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A Bing custom search tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a Bing custom search tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An OpenAPI tool call.

The unique ID of the tool call generated by the model.

The name of the OpenAPI operation being called.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an OpenAPI tool call.

The unique ID of the tool call generated by the model.

The name of the OpenAPI operation that was called.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A browser automation tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a browser automation tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A Fabric data agent tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a Fabric data agent tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An Azure Function tool call.

The unique ID of the tool call generated by the model.

The name of the Azure Function being called.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an Azure Function tool call.

The unique ID of the tool call generated by the model.

The name of the Azure Function that was called.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An A2A (Agent-to-Agent) tool call.

The unique ID of the tool call generated by the model.

The name of the A2A agent card being called.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an A2A (Agent-to-Agent) tool call.

The unique ID of the tool call generated by the model.

The name of the A2A agent card that was called.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A memory search tool call.

The status of a tool call.

The results returned from the memory search.

A memory command tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool, including action (remember or forget) and content.

The status of a tool call.

The output of a memory command tool call.

The unique ID of the tool call generated by the model.

The status of a tool call.

An output message from the model.

The unique ID of the output message.

The type of the output message. Always message.

The role of the output message. Always assistant.

The content of the output message.

Labels an assistant message as intermediate commentary (commentary) or the final answer (final_answer). For models like gpt-5.3-codex and beyond, when sending follow-up requests, preserve and resend phase on all assistant messages — dropping it can degrade performance. Not used for user messages.

The status of the message input. One of in_progress, completed, or incomplete. Populated when input items are returned via API.

The results of a file search tool call. See the file search guide for more information.

The unique ID of the file search tool call.

The type of the file search tool call. Always file_search_call.

The status of the file search tool call. One of in_progress, searching, incomplete or failed,

The queries used to search for files.

A tool call to run a function. See the function calling guide for more information.

The unique ID of the function tool call.

The type of the function tool call. Always function_call.

The unique ID of the function tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

The namespace of the function to run.

The name of the function to run.

A JSON string of the arguments to pass to the function.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

Whether the function tool call runs asynchronously.

The output of a function tool call.

The unique ID of the function tool call output. Populated when this item is returned via API.

The type of the function tool call output. Always function_call_output.

The unique ID of the function tool call generated by the model.

The name of the tool that produced the output.

The namespace of the tool that produced the output.

The execution context that produced this tool call.

The caller type. Always direct.

The caller type. Always program.

The call ID of the program item that produced this tool call.

The output from the function call generated by your code. Can be a string or an list of output content.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

The results of a web search tool call. See the web search guide for more information.

The unique ID of the web search tool call.

The type of the web search tool call. Always web_search_call.

An object describing the specific action taken in this web search call. Includes details on how the model used the web (search, open_page, find_in_page).

The action type.

The search query.

The search queries.

The sources used in the search.

The action type.

The URL opened by the model.

The action type.

The URL of the page searched for the pattern, when available.

The pattern or text to search for within the page.

A tool call to a computer use tool. See the computer use guide for more information.

The type of the computer call. Always computer_call.

The unique ID of the computer call.

An identifier used when responding to the tool call with output.

A click action.

Specifies the event type. For a click action, this property is always click.

A double click action.

Specifies the event type. For a double click action, this property is always set to double_click.

A drag action.

Specifies the event type. For a drag action, this property is always set to drag.

An array of coordinates representing the path of the drag action. Coordinates will appear as an array of objects, eg [ { x: 100, y: 200 }, { x: 200, y: 300 } ]

A collection of keypresses the model would like to perform.

Specifies the event type. For a keypress action, this property is always set to keypress.

The combination of keys the model is requesting to be pressed. This is an array of strings, each representing a key.

A mouse move action.

Specifies the event type. For a move action, this property is always set to move.

A screenshot action.

Specifies the event type. For a screenshot action, this property is always set to screenshot.

A scroll action.

Specifies the event type. For a scroll action, this property is always set to scroll.

An action to type in text.

Specifies the event type. For a type action, this property is always set to type.

The text to type.

A wait action.

Specifies the event type. For a wait action, this property is always set to wait.

Flattened batched actions for computer_use. Each action includes an type discriminator and action-specific fields.

The pending safety checks for the computer call.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

The output of a computer tool call.

The type of the computer tool call output. Always computer_call_output.

The ID of the computer tool call output.

The ID of the computer tool call that produced the output.

The safety checks reported by the API that have been acknowledged by the developer.

A computer screenshot image used with the computer use tool.

Specifies the event type. For a computer screenshot, this property is always set to computer_screenshot.

The URL of the screenshot image.

The identifier of an uploaded file that contains the screenshot.

The status of the message input. One of in_progress, completed, or incomplete. Populated when input items are returned via API.

A description of the chain of thought used by a reasoning model while generating a response. Be sure to include these items in your input to the Responses API for subsequent turns of a conversation if you are manually managing context.

The type of the object. Always reasoning.

The unique identifier of the reasoning content.

Reasoning summary content.

Reasoning text content.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

The type of the item. Always program.

The unique ID of the program item.

The stable call ID of the program item.

The JavaScript source executed by programmatic tool calling.

Opaque program replay fingerprint that must be round-tripped.

The type of the item. Always program_output.

The unique ID of the program output item.

The call ID of the program item.

The result produced by the program item.

The type of the item. Always tool_search_call.

The unique ID of the tool search call item.

Arguments used for the tool search call.

The identifier of the actor that created the item.

The type of the item. Always tool_search_output.

The unique ID of the tool search output item.

The loaded tool definitions returned by tool search.

The identifier of the actor that created the item.

The type of the item. Always additional_tools.

The unique ID of the additional tools item.

The additional tool definitions made available at this item.

A compaction item generated by the v1/responses/compact API.

The type of the item. Always compaction.

The unique ID of the compaction item.

The encrypted content that was produced by compaction.

The identifier of the actor that created the item.

An image generation request made by the model.

The type of the image generation call. Always image_generation_call.

The unique ID of the image generation call.

The status of the image generation call.

The error type.

A description of the error.

The stage where moderation blocked the image generation call.

Public moderation categories. New category values may be added over time.

A tool call to run code.

The type of the code interpreter tool call. Always code_interpreter_call.

The unique ID of the code interpreter tool call.

The status of the code interpreter tool call. Valid values are in_progress, completed, incomplete, interpreting, and failed.

The ID of the container used to run the code.

A tool call to run a command on the local shell.

The type of the local shell call. Always local_shell_call.

The unique ID of the local shell call.

The unique ID of the local shell tool call generated by the model.

Execute a shell command on the server.

The type of the local shell action. Always exec.

The command to run.

Environment variables to set for the command.

The status of the local shell call.

The output of a local shell tool call.

The type of the local shell tool call output. Always local_shell_call_output.

The unique ID of the local shell tool call generated by the model.

A JSON string of the output of the local shell tool call.

A tool call that executes one or more shell commands in a managed environment.

The type of the item. Always shell_call.

The unique ID of the shell tool call. Populated when this item is returned via API.

The unique ID of the shell tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

Execute a shell command.

Represents the use of a local environment to perform shell actions.

The environment type. Always local.

Represents a container created with /v1/containers.

The environment type. Always container_reference.

The ID of the entity that created this tool call.

The output of a shell tool call that was emitted.

The type of the shell call output. Always shell_call_output.

The unique ID of the shell call output. Populated when this item is returned via API.

The unique ID of the shell tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

An array of shell call output contents

The identifier of the actor that created the item.

A tool call that applies file diffs by creating, deleting, or updating files.

The type of the item. Always apply_patch_call.

The unique ID of the apply patch tool call. Populated when this item is returned via API.

The unique ID of the apply patch tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

One of the create_file, delete_file, or update_file operations applied via apply_patch.

Instruction describing how to create a file via the apply_patch tool.

Create a new file with the provided diff.

Path of the file to create.

Diff to apply.

Instruction describing how to delete a file via the apply_patch tool.

Delete the specified file.

Path of the file to delete.

Instruction describing how to update a file via the apply_patch tool.

Update an existing file with the provided diff.

Path of the file to update.

Diff to apply.

The ID of the entity that created this tool call.

The output emitted by an apply patch tool call.

The type of the item. Always apply_patch_call_output.

The unique ID of the apply patch tool call output. Populated when this item is returned via API.

The unique ID of the apply patch tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

The ID of the entity that created this tool call output.

An invocation of a tool on an MCP server.

The type of the item. Always mcp_call.

The unique ID of the tool call.

The label of the MCP server running the tool.

The name of the tool that was run.

A JSON string of the arguments passed to the tool.

The error from the tool call, if any.

A list of tools available on an MCP server.

The type of the item. Always mcp_list_tools.

The unique ID of the list.

The label of the MCP server.

The tools available on the server.

A request for human approval of a tool invocation.

The type of the item. Always mcp_approval_request.

The unique ID of the approval request.

The label of the MCP server making the request.

The name of the tool to run.

A JSON string of arguments for the tool.

A response to an MCP approval request.

The type of the item. Always mcp_approval_response.

The unique ID of the approval response

The ID of the approval request being answered.

Whether the request was approved.

The type of the custom tool call. Always custom_tool_call.

The unique ID of the custom tool call in the OpenAI platform.

An identifier used to map this custom tool call to a tool call output.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

The namespace of the custom tool being called.

The name of the custom tool being called.

The input for the custom tool call generated by the model.

Whether the custom tool call runs asynchronously.

The identifier of the actor that created the item.

The type of the custom tool call output. Always custom_tool_call_output.

The unique ID of the custom tool call output in the OpenAI platform.

The call ID, used to map this custom tool call output to a custom tool call.

The execution context that produced this tool call.

The caller type. Always direct.

The caller type. Always program.

The call ID of the program item that produced this tool call.

The output from the custom tool call generated by your code. Can be a string or an list of output content.

The identifier of the actor that created the item.

The type of the event. Always response.reasoning_summary_part.added.

The ID of the item this summary part is associated with.

The type of the event. Always response.reasoning_summary_text.delta.

The ID of the item this summary text delta is associated with.

The text delta that was added to the summary.

The type of the event. Always response.reasoning_text.delta.

The ID of the item this reasoning text delta is associated with.

The text delta that was added to the reasoning content.

The type of the event. Always response.refusal.delta.

The ID of the output item that the refusal text is added to.

The refusal text that is added.

The type of the event. Always response.output_text.delta.

The ID of the output item that the text delta was added to.

The text delta that was added.

The log probabilities of the tokens in the delta.

The type of the event. Always response.web_search_call.in_progress.

Unique ID for the output item associated with the web search call.

The type of the event. Always response.web_search_call.searching.

Unique ID for the output item associated with the web search call.

The type of the event. Always 'response.image_generation_call.generating'.

The unique identifier of the image generation item being processed.

The type of the event. Always 'response.image_generation_call.in_progress'.

The unique identifier of the image generation item being processed.

The type of the event. Always 'response.image_generation_call.partial_image'.

The unique identifier of the image generation item being processed.

Base64-encoded partial image data, suitable for rendering as an image.

The image size that was used.

The image quality that was used.

The background setting that was used.

The output format that was used.

The type of the event. Always 'response.mcp_call_arguments.delta'.

The unique identifier of the MCP tool call item being processed.

A JSON string containing the partial update to the arguments for the MCP tool call.

The type of the event. Always 'response.mcp_call.failed'.

The ID of the MCP tool call item that failed.

The type of the event. Always 'response.mcp_call.in_progress'.

The unique identifier of the MCP tool call item being processed.

The type of the event. Always 'response.mcp_list_tools.failed'.

The ID of the MCP tool call item that failed.

The type of the event. Always 'response.mcp_list_tools.in_progress'.

The ID of the MCP tool call item that is being processed.

The type of the event. Always 'response.output_text.annotation.added'.

The unique identifier of the item to which the annotation is being added.

An annotation that applies to a span of output text.

A citation to a file.

The type of the file citation. Always file_citation.

The ID of the file.

The filename of the file cited.

A citation for a web resource used to generate a model response.

The type of the URL citation. Always url_citation.

The URL of the web resource.

The title of the web resource.

A citation for a container file used to generate a model response.

The type of the container file citation. Always container_file_citation.

The ID of the container file.

The ID of the file.

The filename of the container file cited.

A path to a file.

The type of the file path. Always file_path.

The ID of the file.

The type of the event. Always 'response.queued'.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Unique identifier for this Response.

The object type of this resource - always set to response.

The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete.

Unix timestamp (in seconds) of when this Response was created.

An error object returned when the model fails to generate a Response.

The original error type, or null if it was not recorded.

The request parameter associated with the error, or null if none was recorded.

The error code for the response.

A human-readable description of the error.

The public explanation for this block.

The public continuation instruction.

An array of content items generated by the model. - The length and order of items in the output array is dependent on the model's response. - Rather than accessing the first item in the output array and assuming it's an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.

Configuration options for reasoning models.

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.

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

The prompt-caching options that were applied to the response. Supported for gpt-5.6 and later models.

Prompt cache diagnostics requested for this response.

Moderation results or errors for the response input and output.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Whether to allow the model to run tool calls in parallel.

The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation.

The unique ID of the conversation that this response was associated with.

The agent used for this response

The name of the agent.

The version identifier of the agent.

Model selection metadata providing visibility into routing decisions.

Routing decision metadata from Model Router.

The routing mode used by Model Router for the request.

Effective Model Router session affinity metadata for a request.

Effective modes reported for Model Router session affinity.

Conversation identifier sources supported by Model Router session affinity.

Final outcomes reported for Model Router session affinity.

Ordered trace of routing decisions. Contains one entry for Chat Completions, and may contain multiple entries for Responses API or agent scenarios where multiple inference turns occur.

The content filter evaluation results.

The event type identifier.

Unique identifier for the API item associated with this event.

The incremental input data (delta) for the custom tool call.

The type of the event. Always response.audio.done.

The type of the event. Always response.audio.transcript.done.

The type of the event. Always response.code_interpreter_call_code.done.

The unique identifier of the code interpreter tool call item.

The final code snippet output by the code interpreter.

The type of the event. Always response.code_interpreter_call.completed.

The unique identifier of the code interpreter tool call item.

The type of the event. Always response.completed.

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.

This field is being replaced by safety_identifier and prompt_cache_key. Use prompt_cache_key instead to maintain caching optimizations. A stable identifier for your end-users. Used to boost cache hit rates by better bucketing similar requests and to help OpenAI detect and prevent abuse. Learn more.

The model deployment to use for the creation of this response.

Configuration options for a text response from the model. Can be plain text or structured JSON data. Learn more: - Text inputs and outputs - Structured Outputs

An object specifying the format that the model must output. Configuring { "type": "json_schema" } enables Structured Outputs, which ensures the model will match your supplied JSON schema. Learn more in the Structured Outputs guide. The default format is { "type": "text" } with no additional options. Not recommended for gpt-4o and newer models:* Setting to { "type": "json_object" } enables the older JSON mode, which ensures the message the model generates is valid JSON. Using json_schema is preferred for models that support it.

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

The type of response format being defined. Always json_schema.

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

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.

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

Default response format. Used to generate text responses.

The type of response format being defined. Always text.

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.

The type of response format being defined. Always json_object.

Constrains the verbosity of the model's response. Lower values will result in more concise responses, while higher values will result in more verbose responses. Currently supported values are low, medium, and high. The default is medium.

An array of tools the model may call while generating a response. You can specify which tool to use by setting the tool_choice parameter. We support the following categories of tools: - Built-in tools: Tools that are provided by OpenAI that extend the model's capabilities, like web search or file search. Learn more about built-in tools. - MCP Tools: Integrations with third-party systems via custom MCP servers or predefined connectors such as Google Drive and SharePoint. Learn more about MCP Tools. - Function calls (custom tools): Functions that are defined by you, enabling the model to call your own code with strongly typed arguments and outputs. Learn more about function calling. You can also use custom tools to call your own code.

Controls which (if any) tool is called by the model. none means the model will not call any tool 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.

Use this option to force the model to call a specific function.

For function calling, the type is always function.

The name of the function to call.

Use this option to force the model to call a specific tool on a remote MCP server.

For MCP tools, the type is always mcp.

The label of the MCP server to use.

Constrains the tools available to the model to a pre-defined set.

Allowed tool configuration type. Always allowed_tools.

Constrains the tools available to the model to a pre-defined set. auto allows the model to pick from among the allowed tools and generate a message. required requires the model to call one or more of the allowed tools.

A list of tool definitions that the model should be allowed to call. For the Responses API, the list of tool definitions might look like: json [ { "type": "function", "name": "get_weather" }, { "type": "mcp", "server_label": "deepwiki" }, { "type": "image_generation" } ]

Use this option to force the model to call a specific custom tool.

For custom tool calling, the type is always custom.

The name of the custom tool to call.

The tool to call. Always programmatic_tool_calling.

Forces the model to call the apply_patch tool when executing a tool call.

The tool to call. Always apply_patch.

Forces the model to call the shell tool when a tool call is required.

The tool to call. Always shell.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Indicates that the model should use a built-in tool to generate a response. Learn more about built-in tools.

Reference to a prompt template and its variables. Learn more.

The unique identifier of the prompt template to use.

Optional map of values to substitute in for variables in your prompt. The substitution values can either be strings, or other Response input types like images or files.

The type of the input item. Always input_text.

The text input to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_image.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The type of the input item. Always input_file.

The name of the file to be sent to the model.

The content of the file to be sent to the model.

Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

The breakpoint mode. Always explicit.

The URL of the file to be sent to the model.

Specifies the processing type used for serving the request. - If set to 'auto', then the request will be processed with the service tier configured in the Project settings. Unless otherwise configured, the Project will use 'default'. - If set to 'default', then the request will be processed with the standard pricing and performance for the selected model. - If set to 'flex', then the request will be processed with the Flex Processing service tier. - To opt-in to Fast mode at the request level, include the service_tier=fast or service_tier=priority parameter for Responses or Chat Completions. The response will show service_tier=priority regardless of if you specify service_tier=fast or priority in your request. - If set to 'ultrafast', then the request will be processed with the access-controlled Ultrafast Processing service tier. This tier is currently available for gpt-5.6-sol; a response served through it will show service_tier=ultrafast. - When not set, the default behavior is 'auto'. When the service_tier parameter is set, the response body will include the service_tier value based on the processing mode actually used to serve the request. This response value may be different from the value set in the parameter.

Unique identifier for this Response.

The object type of this resource - always set to response.

The status of the response generation. One of completed, failed, in_progress, cancelled, queued, or incomplete.

Unix timestamp (in seconds) of when this Response was created.

An error object returned when the model fails to generate a Response.

The original error type, or null if it was not recorded.

The request parameter associated with the error, or null if none was recorded.

The error code for the response.

A human-readable description of the error.

The public explanation for this block.

The public continuation instruction.

An array of content items generated by the model. - The length and order of items in the output array is dependent on the model's response. - Rather than accessing the first item in the output array and assuming it's an assistant message with the content generated by the model, you might consider using the output_text property where supported in SDKs.

Configuration options for reasoning models.

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.

Represents token usage details including input tokens, output tokens, a breakdown of output tokens, and the total tokens used.

The prompt-caching options that were applied to the response. Supported for gpt-5.6 and later models.

Prompt cache diagnostics requested for this response.

Moderation results or errors for the response input and output.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Moderation results or an error for a response moderation check.

An error produced while attempting moderation for the response input or output.

The object type, which was always error for moderation failures.

The error code.

The error message.

A moderation result produced for the response input or output.

The object type, which was always moderation_result for successful moderation results.

The moderation model that produced this result.

A boolean indicating whether the content was flagged by any category.

A dictionary of moderation categories to booleans, True if the input is flagged under this category.

A dictionary of moderation categories to scores.

Which modalities of input are reflected by the score for each category.

Whether to allow the model to run tool calls in parallel.

The conversation that this response belonged to. Input items and output items from this response were automatically added to this conversation.

The unique ID of the conversation that this response was associated with.

The agent used for this response

The name of the agent.

The version identifier of the agent.

Model selection metadata providing visibility into routing decisions.

Routing decision metadata from Model Router.

The routing mode used by Model Router for the request.

Effective Model Router session affinity metadata for a request.

Effective modes reported for Model Router session affinity.

Conversation identifier sources supported by Model Router session affinity.

Final outcomes reported for Model Router session affinity.

Ordered trace of routing decisions. Contains one entry for Chat Completions, and may contain multiple entries for Responses API or agent scenarios where multiple inference turns occur.

The content filter evaluation results.

The type of the event. Always response.content_part.done.

The ID of the output item that the content part was added to.

A text output from the model.

The type of the output text. Always output_text.

The text output from the model.

The annotations of the text output.

A refusal from the model.

The type of the refusal. Always refusal.

The refusal explanation from the model.

Reasoning text from the model.

The type of the reasoning text. Always reasoning_text.

The reasoning text from the model.

The type of the event. Always response.file_search_call.completed.

The ID of the output item that the file search call is initiated.

The ID of the item.

The function-call arguments.

The type of the event. Always response.output_item.done.

The structured output captured during the response.

The kind of CSDL action (e.g., 'SetVariable', 'InvokeAzureAgent').

Unique identifier for the action.

ID of the parent action if this is a nested action.

ID of the previous action if this action follows another.

Status of the action (e.g., 'in_progress', 'completed', 'failed', 'cancelled').

Request from the service for the user to perform OAuth consent.

The link the user can use to perform OAuth consent.

The server label for the OAuth consent request.

A Bing grounding tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a Bing grounding tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A SharePoint grounding tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a SharePoint grounding tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An Azure AI Search tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an Azure AI Search tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A Bing custom search tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a Bing custom search tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An OpenAPI tool call.

The unique ID of the tool call generated by the model.

The name of the OpenAPI operation being called.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an OpenAPI tool call.

The unique ID of the tool call generated by the model.

The name of the OpenAPI operation that was called.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A browser automation tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a browser automation tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A Fabric data agent tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of a Fabric data agent tool call.

The unique ID of the tool call generated by the model.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An Azure Function tool call.

The unique ID of the tool call generated by the model.

The name of the Azure Function being called.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an Azure Function tool call.

The unique ID of the tool call generated by the model.

The name of the Azure Function that was called.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

An A2A (Agent-to-Agent) tool call.

The unique ID of the tool call generated by the model.

The name of the A2A agent card being called.

A JSON string of the arguments to pass to the tool.

The status of a tool call.

The output of an A2A (Agent-to-Agent) tool call.

The unique ID of the tool call generated by the model.

The name of the A2A agent card that was called.

The output content from a tool call, which can be a dictionary, string, or array.

The status of a tool call.

A memory search tool call.

The status of a tool call.

The results returned from the memory search.

A memory command tool call.

The unique ID of the tool call generated by the model.

A JSON string of the arguments to pass to the tool, including action (remember or forget) and content.

The status of a tool call.

The output of a memory command tool call.

The unique ID of the tool call generated by the model.

The status of a tool call.

An output message from the model.

The unique ID of the output message.

The type of the output message. Always message.

The role of the output message. Always assistant.

The content of the output message.

Labels an assistant message as intermediate commentary (commentary) or the final answer (final_answer). For models like gpt-5.3-codex and beyond, when sending follow-up requests, preserve and resend phase on all assistant messages — dropping it can degrade performance. Not used for user messages.

The status of the message input. One of in_progress, completed, or incomplete. Populated when input items are returned via API.

The results of a file search tool call. See the file search guide for more information.

The unique ID of the file search tool call.

The type of the file search tool call. Always file_search_call.

The status of the file search tool call. One of in_progress, searching, incomplete or failed,

The queries used to search for files.

A tool call to run a function. See the function calling guide for more information.

The unique ID of the function tool call.

The type of the function tool call. Always function_call.

The unique ID of the function tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

The namespace of the function to run.

The name of the function to run.

A JSON string of the arguments to pass to the function.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

Whether the function tool call runs asynchronously.

The output of a function tool call.

The unique ID of the function tool call output. Populated when this item is returned via API.

The type of the function tool call output. Always function_call_output.

The unique ID of the function tool call generated by the model.

The name of the tool that produced the output.

The namespace of the tool that produced the output.

The execution context that produced this tool call.

The caller type. Always direct.

The caller type. Always program.

The call ID of the program item that produced this tool call.

The output from the function call generated by your code. Can be a string or an list of output content.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

The results of a web search tool call. See the web search guide for more information.

The unique ID of the web search tool call.

The type of the web search tool call. Always web_search_call.

An object describing the specific action taken in this web search call. Includes details on how the model used the web (search, open_page, find_in_page).

The action type.

The search query.

The search queries.

The sources used in the search.

The action type.

The URL opened by the model.

The action type.

The URL of the page searched for the pattern, when available.

The pattern or text to search for within the page.

A tool call to a computer use tool. See the computer use guide for more information.

The type of the computer call. Always computer_call.

The unique ID of the computer call.

An identifier used when responding to the tool call with output.

A click action.

Specifies the event type. For a click action, this property is always click.

A double click action.

Specifies the event type. For a double click action, this property is always set to double_click.

A drag action.

Specifies the event type. For a drag action, this property is always set to drag.

An array of coordinates representing the path of the drag action. Coordinates will appear as an array of objects, eg [ { x: 100, y: 200 }, { x: 200, y: 300 } ]

A collection of keypresses the model would like to perform.

Specifies the event type. For a keypress action, this property is always set to keypress.

The combination of keys the model is requesting to be pressed. This is an array of strings, each representing a key.

A mouse move action.

Specifies the event type. For a move action, this property is always set to move.

A screenshot action.

Specifies the event type. For a screenshot action, this property is always set to screenshot.

A scroll action.

Specifies the event type. For a scroll action, this property is always set to scroll.

An action to type in text.

Specifies the event type. For a type action, this property is always set to type.

The text to type.

A wait action.

Specifies the event type. For a wait action, this property is always set to wait.

Flattened batched actions for computer_use. Each action includes an type discriminator and action-specific fields.

The pending safety checks for the computer call.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

The output of a computer tool call.

The type of the computer tool call output. Always computer_call_output.

The ID of the computer tool call output.

The ID of the computer tool call that produced the output.

The safety checks reported by the API that have been acknowledged by the developer.

A computer screenshot image used with the computer use tool.

Specifies the event type. For a computer screenshot, this property is always set to computer_screenshot.

The URL of the screenshot image.

The identifier of an uploaded file that contains the screenshot.

The status of the message input. One of in_progress, completed, or incomplete. Populated when input items are returned via API.

A description of the chain of thought used by a reasoning model while generating a response. Be sure to include these items in your input to the Responses API for subsequent turns of a conversation if you are manually managing context.

The type of the object. Always reasoning.

The unique identifier of the reasoning content.

Reasoning summary content.

Reasoning text content.

The status of the item. One of in_progress, completed, or incomplete. Populated when items are returned via API.

The type of the item. Always program.

The unique ID of the program item.

The stable call ID of the program item.

The JavaScript source executed by programmatic tool calling.

Opaque program replay fingerprint that must be round-tripped.

The type of the item. Always program_output.

The unique ID of the program output item.

The call ID of the program item.

The result produced by the program item.

The type of the item. Always tool_search_call.

The unique ID of the tool search call item.

Arguments used for the tool search call.

The identifier of the actor that created the item.

The type of the item. Always tool_search_output.

The unique ID of the tool search output item.

The loaded tool definitions returned by tool search.

The identifier of the actor that created the item.

The type of the item. Always additional_tools.

The unique ID of the additional tools item.

The additional tool definitions made available at this item.

A compaction item generated by the v1/responses/compact API.

The type of the item. Always compaction.

The unique ID of the compaction item.

The encrypted content that was produced by compaction.

The identifier of the actor that created the item.

An image generation request made by the model.

The type of the image generation call. Always image_generation_call.

The unique ID of the image generation call.

The status of the image generation call.

The error type.

A description of the error.

The stage where moderation blocked the image generation call.

Public moderation categories. New category values may be added over time.

A tool call to run code.

The type of the code interpreter tool call. Always code_interpreter_call.

The unique ID of the code interpreter tool call.

The status of the code interpreter tool call. Valid values are in_progress, completed, incomplete, interpreting, and failed.

The ID of the container used to run the code.

A tool call to run a command on the local shell.

The type of the local shell call. Always local_shell_call.

The unique ID of the local shell call.

The unique ID of the local shell tool call generated by the model.

Execute a shell command on the server.

The type of the local shell action. Always exec.

The command to run.

Environment variables to set for the command.

The status of the local shell call.

The output of a local shell tool call.

The type of the local shell tool call output. Always local_shell_call_output.

The unique ID of the local shell tool call generated by the model.

A JSON string of the output of the local shell tool call.

A tool call that executes one or more shell commands in a managed environment.

The type of the item. Always shell_call.

The unique ID of the shell tool call. Populated when this item is returned via API.

The unique ID of the shell tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

Execute a shell command.

Represents the use of a local environment to perform shell actions.

The environment type. Always local.

Represents a container created with /v1/containers.

The environment type. Always container_reference.

The ID of the entity that created this tool call.

The output of a shell tool call that was emitted.

The type of the shell call output. Always shell_call_output.

The unique ID of the shell call output. Populated when this item is returned via API.

The unique ID of the shell tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

An array of shell call output contents

The identifier of the actor that created the item.

A tool call that applies file diffs by creating, deleting, or updating files.

The type of the item. Always apply_patch_call.

The unique ID of the apply patch tool call. Populated when this item is returned via API.

The unique ID of the apply patch tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

One of the create_file, delete_file, or update_file operations applied via apply_patch.

Instruction describing how to create a file via the apply_patch tool.

Create a new file with the provided diff.

Path of the file to create.

Diff to apply.

Instruction describing how to delete a file via the apply_patch tool.

Delete the specified file.

Path of the file to delete.

Instruction describing how to update a file via the apply_patch tool.

Update an existing file with the provided diff.

Path of the file to update.

Diff to apply.

The ID of the entity that created this tool call.

The output emitted by an apply patch tool call.

The type of the item. Always apply_patch_call_output.

The unique ID of the apply patch tool call output. Populated when this item is returned via API.

The unique ID of the apply patch tool call generated by the model.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

The ID of the entity that created this tool call output.

An invocation of a tool on an MCP server.

The type of the item. Always mcp_call.

The unique ID of the tool call.

The label of the MCP server running the tool.

The name of the tool that was run.

A JSON string of the arguments passed to the tool.

The error from the tool call, if any.

A list of tools available on an MCP server.

The type of the item. Always mcp_list_tools.

The unique ID of the list.

The label of the MCP server.

The tools available on the server.

A request for human approval of a tool invocation.

The type of the item. Always mcp_approval_request.

The unique ID of the approval request.

The label of the MCP server making the request.

The name of the tool to run.

A JSON string of arguments for the tool.

A response to an MCP approval request.

The type of the item. Always mcp_approval_response.

The unique ID of the approval response

The ID of the approval request being answered.

Whether the request was approved.

The type of the custom tool call. Always custom_tool_call.

The unique ID of the custom tool call in the OpenAI platform.

An identifier used to map this custom tool call to a tool call output.

The execution context that produced this tool call.

The call ID of the program item that produced this tool call.

The namespace of the custom tool being called.

The name of the custom tool being called.

The input for the custom tool call generated by the model.

Whether the custom tool call runs asynchronously.

The identifier of the actor that created the item.

The type of the custom tool call output. Always custom_tool_call_output.

The unique ID of the custom tool call output in the OpenAI platform.

The call ID, used to map this custom tool call output to a custom tool call.

The execution context that produced this tool call.

The caller type. Always direct.

The caller type. Always program.

The call ID of the program item that produced this tool call.

The output from the custom tool call generated by your code. Can be a string or an list of output content.

The identifier of the actor that created the item.

The type of the event. Always response.reasoning_summary_part.done.

The ID of the item this summary part is associated with.

The completion status of the summary part. Omitted when the part completed normally and set to incomplete when generation was interrupted.

The type of the event. Always response.reasoning_summary_text.done.

The ID of the item this summary text is associated with.

The full text of the completed reasoning summary.

The type of the event. Always response.reasoning_text.done.

The ID of the item this reasoning text is associated with.

The full text of the completed reasoning content.

The type of the event. Always response.refusal.done.

The ID of the output item that the refusal text is finalized.

The refusal text that is finalized.

The type of the event. Always response.output_text.done.

The ID of the output item that the text content is finalized.

The text content that is finalized.

The log probabilities of the tokens in the delta.

The type of the event. Always response.web_search_call.completed.

Unique ID for the output item associated with the web search call.

The type of the event. Always 'response.image_generation_call.completed'.

The unique identifier of the image generation item being processed.

The type of the event. Always 'response.mcp_call_arguments.done'.

The unique identifier of the MCP tool call item being processed.

A JSON string containing the finalized arguments for the MCP tool call.

The type of the event. Always 'response.mcp_call.completed'.

The ID of the MCP tool call item that completed.

The type of the event. Always 'response.mcp_list_tools.completed'.

The ID of the MCP tool call item that produced this output.

The event type identifier.

Unique identifier for the API item associated with this event.

The complete input data for the custom tool call.

The public explanation for this block.

The public continuation instruction.

The public explanation for this block.

The public continuation instruction.