> ## Documentation Index > Fetch the complete documentation index at: https://sambanova-systems.mintlify.site/docs/llms.txt > Use this file to discover all available pages before exploring further. # Create a model response > Creates a model response for the given input. Only `type: "function"` tools are supported; other tool types are filtered server-side. SambaNova is stateless, conversation history must be supplied in full via `input[]` on each request. ## OpenAPI ````yaml https://app.stainless.com/api/spec/documented/sambanova/openapi.documented.yml post /responses openapi: 3.1.1 info: title: SambaNova cloud API description: SambaNova cloud API Specification version: 1.2.0 termsOfService: https://sambanova.ai/cloud-end-user-license-agreement contact: email: info@sambanova.ai name: SambaNova information license: name: Apache 2.0 url: https://www.apache.org/licenses/LICENSE-2.0.html servers: - url: https://api.sambanova.ai/v1 security: - api_key: [] externalDocs: description: Find out more in the official SambaNova docs url: https://docs.sambanova.ai/docs/en/api-reference/overview paths: /responses: post: tags: - Responses summary: Create a model response description: >- Creates a model response for the given input. Only `type: "function"` tools are supported; other tool types are filtered server-side. SambaNova is stateless, conversation history must be supplied in full via `input[]` on each request. operationId: createResponse requestBody: required: true description: Response creation parameters content: application/json: schema: $ref: '#/components/schemas/ResponseRequest' responses: '200': description: >- Successful response. Returns a ResponseResponse object (non-streaming), or a stream of server-sent ResponseStreamEvent object events ending with a response.completed event (when stream: true). content: application/json: schema: oneOf: - $ref: '#/components/schemas/ResponseResponse' - $ref: '#/components/schemas/ResponseStreamEvent' '400': description: Bad Request — missing or invalid parameters content: application/json: schema: oneOf: - $ref: '#/components/schemas/GeneralError' - $ref: '#/components/schemas/SimpleError' examples: missing_input: summary: Required field missing value: error: message: 'Missing required field: ''input''.' type: invalid_request_error param: input code: missing_required_field request_id: abc123 invalid_role: summary: Invalid role on input item value: error: message: >- Invalid value: 'foo'. Supported values are: 'user', 'assistant', 'system', 'developer'. type: invalid_request_error param: input[1].role code: invalid_value request_id: abc124 empty_input: summary: Empty input array value: error: message: >- Invalid value for 'input': expected a non-empty array or string. type: invalid_request_error param: input code: invalid_value request_id: abc125 invalid_json: summary: Malformed JSON body value: error: code: null message: >- We could not parse the JSON body of your request. (HINT: This likely means you aren't using your HTTP library correctly. A JSON payload is expected, but what was sent was not valid JSON.) param: null type: invalid_request_error request_id: 3f7db127 adapter_mapping_failed: summary: Server failed to parse model tool call output value: error: code: adapter_mapping_failed error_model_output: |- { "name": "get_weather", "arguments": { city: "Bogotá" } } message: >- Failed to parse tool call from GPT OSS output: Expecting property name enclosed in double quotes: line 4 column 5 (char 48) param: null type: server_error request_id: 494e734dc82b4b37bd914238c79c3e6c simple: summary: Simple error (unhandled cases) value: error: Unhandled error '401': description: Unauthorized — invalid or missing API key content: application/json: schema: $ref: '#/components/schemas/GeneralError' examples: invalid_api_key: summary: Invalid API key value: error: message: >- Incorrect API key provided: *****. You can find your API key at https://cloud.sambanova.ai/apis. type: invalid_request_error param: null code: invalid_api_key request_id: abc126 missing_api_key: summary: No API key provided value: error: message: >- You didn't provide an API key. You need to provide your API key in an Authorization header using Bearer auth (i.e. Authorization: Bearer YOUR_KEY). type: invalid_request_error param: null code: null request_id: abc127 '404': description: Not found — model does not exist or is not accessible content: application/json: schema: $ref: '#/components/schemas/GeneralError' examples: model_not_found: summary: Model does not exist or is not accessible value: error: message: >- The model `abc` does not exist or you do not have access to it. type: invalid_request_error param: model code: model_not_found request_id: abc128 simple: summary: Simple error (unhandled cases e.g. wrong endpoint) value: error: Not found '408': description: Request Timeout content: application/json: schema: $ref: '#/components/schemas/SimpleError' '410': description: Gone — model is no longer available (deprecated or removed) content: application/json: schema: $ref: '#/components/schemas/SimpleError' '429': description: Too Many Requests — rate limit exceeded content: application/json: schema: $ref: '#/components/schemas/GeneralError' '500': description: Internal Server Error — unexpected issue on server side content: text/plain: schema: type: string '503': description: Service Temporarily Unavailable content: text/plain: schema: type: string example: Service Temporarily Unavailable security: - api_key: [] x-codeSamples: - lang: JavaScript source: |- import SambaNova from 'sambanova'; const client = new SambaNova({ apiKey: process.env['SAMBANOVA_API_KEY'], // This is the default and can be omitted }); const response = await client.responses.create({ input: [ { content: 'What is the weather in San Francisco?', role: 'user', type: 'message', }, { content: [ { text: 'The weather in San Francisco is 65°F and partly cloudy.', type: 'output_text' }, ], role: 'assistant', type: 'message', }, { content: 'What should I wear?', role: 'user', type: 'message', }, ], model: 'gpt-oss-120b', }); console.log(response); - lang: Python source: |- import os from sambanova import SambaNova client = SambaNova( api_key=os.environ.get("SAMBANOVA_API_KEY"), # This is the default and can be omitted ) for response in client.responses.create( input=[{ "content": "What is the weather in San Francisco?", "role": "user", "type": "message", }, { "content": [{ "text": "The weather in San Francisco is 65°F and partly cloudy.", "type": "output_text", }], "role": "assistant", "type": "message", }, { "content": "What should I wear?", "role": "user", "type": "message", }], model="gpt-oss-120b", ): print(response) components: schemas: ResponseRequest: title: Response Request type: object description: responses request object additionalProperties: true properties: model: title: Model description: >- The model ID to use (e.g. gpt-oss-120b). See available [models](https://docs.sambanova.ai/docs/en/models/sambacloud-models) anyOf: - type: string - enum: - gpt-oss-120b - MiniMax-M2.5 - MiniMax-M2.7 input: title: Input description: >- Text input to the model, or a structured list of input items representing the full conversation turn. A plain string is equivalent to a single user message. oneOf: - type: string title: Text input description: A plain text input equivalent to a user-role message. - type: array title: Input item list description: >- An ordered list of input items (messages, tool calls, tool results, etc.). items: $ref: '#/components/schemas/ResponseInputItem' examples: - - role: user content: What is the weather in San Francisco? - role: assistant content: - type: output_text text: >- The weather in San Francisco is 65°F and partly cloudy. - role: user content: What should I wear? - - role: user content: What is the weather in San Francisco? - type: function_call id: fc_abc123 call_id: call_abc123 name: get_weather arguments: '{"location":"San Francisco, CA"}' status: completed - type: function_call_output call_id: call_abc123 output: '{"temperature":"65°F","condition":"partly cloudy"}' instructions: title: Instructions type: string description: >- Inserts a system (or developer) message as the first item in the model's context. Equivalent to a system-role message prepended to input[]. nullable: true stream: title: Stream type: boolean description: If true, the response is delivered as server-sent events (SSE). default: false nullable: true max_output_tokens: title: Max Output Tokens type: integer description: >- Upper bound on the number of tokens the model may generate, including visible output tokens and reasoning tokens. nullable: true example: 1024 temperature: title: Temperature type: number description: >- Controls randomness in generation. Range: 0–2. It is recommended to alter this, top_p, or top_k but not more than one at a time. minimum: 0 maximum: 2 default: 0.7 nullable: true example: 0.7 top_p: title: Top P type: number description: >- Nucleus sampling cutoff. Range: 0–1. It is recommended to alter this, temperature, or top_k but not more than one at a time. minimum: 0 maximum: 1 default: 1 nullable: true example: 1 top_k: title: Top K type: integer description: >- Limits sampling to the top K most probable tokens. It is recommended to alter this, top_p, or temperature but not more than one at a time. minimum: 1 maximum: 100 nullable: true example: 5 top_logprobs: title: Top Logprobs description: >- Number of top log-probability entries to return per output token. Null means log probabilities are not returned. type: integer nullable: true frequency_penalty: title: Frequency Penalty type: number description: >- Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim. Not currently implemented; accepted for API compatibility and echoed in the response. minimum: -2 maximum: 2 default: 0 nullable: true presence_penalty: title: Presence Penalty type: number description: >- Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics. Not currently implemented; accepted for API compatibility and echoed in the response. minimum: -2 maximum: 2 default: 0 nullable: true tools: title: Tools type: array description: >- Tools available to the model. Only type: "function" is supported; all other tool types are filtered server-side. items: $ref: '#/components/schemas/ResponseTool' nullable: true maxItems: 128 parallel_tool_calls: title: Parallel Tool Calls type: boolean description: >- Whether the model may issue multiple tool calls in parallel within one turn. default: true nullable: true max_tool_calls: title: Max Tool Calls type: integer description: >- Maximum number of tool calls the model may make in a single response turn. Not currently implemented; accepted for API compatibility. nullable: true tool_choice: $ref: '#/components/schemas/ResponseToolChoiceOption' text: title: Text type: object description: >- Response format configuration. Supports plain text, json_object, and json_schema. properties: format: $ref: '#/components/schemas/ResponseFormatConfiguration' reasoning: title: Reasoning type: object description: >- Reasoning configuration for models that support it. Ignored on non-reasoning models. properties: effort: title: Effort type: string enum: - low - medium - high default: medium nullable: true description: >- Reasoning effort level. "low" is faster with less depth; "high" is deeper at higher token cost. nullable: true user: title: User type: string description: 'Included for API compatibility, but only echoed back in response ' nullable: true metadata: title: Metadata type: object description: Included for API compatibility, but not supported nullable: true additionalProperties: type: string store: title: Store type: boolean description: >- SambaNova is stateless - this field is accepted for API compatibility but has no effect. Always echoed back as false. default: false nullable: true truncation: title: Truncation type: string description: >- Accepted for API compatibility and echoed in the response. Context truncation behavior is not currently configurable via this field in SambaNova. enum: - auto - disabled default: disabled nullable: true previous_response_id: title: Previous Response ID type: string description: >- Not supported. SambaNova is stateless and does not maintain server-side conversation state. Accepted for API compatibility but ignored; clients must supply the full conversation history in input[]. nullable: true background: title: Background type: boolean description: >- Accepted for API compatibility and echoed back in the response. Has no effect on server behavior. nullable: true service_tier: title: Service Tier type: string description: >- Accepted for API compatibility and echoed back in the response. Has no effect on server behavior. nullable: true required: - model - input ResponseResponse: title: Response Response type: object description: >- A response object returned by POST /responses (non-streaming). Contains the model's output items, echoed input parameters, lifecycle metadata, and token usage. properties: id: title: ID type: string description: Unique identifier for this response. object: title: Object type: string description: The object type. Always "response". enum: - response status: title: Status type: string description: >- Lifecycle status of the response. "completed" means the model finished successfully. "failed" means an error occurred during generation. "incomplete" means generation was cut short (e.g. max_output_tokens reached). enum: - completed - failed - in_progress - incomplete created_at: title: Created At type: integer description: Unix timestamp (seconds) when the response was created. completed_at: title: Completed At type: integer description: Unix timestamp (seconds) when the response finished generating. nullable: true model: title: Model type: string description: The model ID used to generate this response. output: title: Output type: array description: >- Ordered array of output items generated by the model. Items may be of type "message", "reasoning", or "function_call". items: $ref: '#/components/schemas/ResponseOutputItem' usage: $ref: '#/components/schemas/ResponseUsage' error: title: Error type: object description: >- In-band error object present when status is "failed". Null when the response completed successfully. properties: code: title: Code type: string description: The error code. message: title: message type: string description: A human-readable error message. nullable: true required: - code - message incomplete_details: title: Incomplete Details type: object description: >- Present when status is "incomplete". Describes why generation stopped before completion (e.g. max_output_tokens reached). properties: reason: type: string description: The reason why the response is incomplete. nullable: true instructions: title: Instructions type: string description: >- The system instructions echoed from the request, or null if none were provided. nullable: true temperature: title: Temperature type: number description: The temperature value used for this response. nullable: true top_p: title: Top P type: number description: The top_p value used for this response. nullable: true top_k: title: Top K type: integer description: The top_k value used for this response. nullable: true frequency_penalty: title: Frequency Penalty type: number description: >- The frequency_penalty value echoed from the request. Not currently implemented; accepted for API compatibility nullable: true presence_penalty: title: Presence Penalty type: number description: >- The presence_penalty value echoed from the request. Not currently implemented; accepted for API compatibility nullable: true tools: title: Tools type: array description: Tool definitions available to the model for this response. items: $ref: '#/components/schemas/ResponseTool' parallel_tool_calls: title: Parallel Tool Calls type: boolean description: Whether parallel tool calls were enabled. max_tool_calls: title: Max Tool Calls type: integer description: The max_tool_calls value echoed from the request. nullable: true tool_choice: $ref: '#/components/schemas/ResponseToolChoiceOption' text: title: Text type: object description: >- The text format configuration (structured output mode) used for this response. properties: format: $ref: '#/components/schemas/ResponseFormatConfiguration' reasoning: title: Reasoning type: object description: The reasoning configuration used for this response. properties: effort: title: Effort type: string description: >- The reasoning effort level that was applied for supported models. enum: - low - medium - high nullable: true summary: title: Summary type: string description: Not supported. nullable: true nullable: true truncation: title: Truncation type: string description: The truncation value echoed from the request. enum: - auto - disabled nullable: true background: title: Background type: boolean description: Whether background generation was requested. nullable: true metadata: title: Metadata type: object description: The metadata echoed from the request. nullable: true additionalProperties: type: string max_output_tokens: title: Max Output Tokens type: integer description: The max_output_tokens limit echoed from the request. nullable: true top_logprobs: title: Top Logprobs type: integer description: The top_logprobs value echoed from the request. nullable: true previous_response_id: title: Previous Response ID type: string description: >- Not supported. Always null. SambaNova is stateless; use input[] to supply full conversation history. nullable: true store: title: Store type: boolean description: >- Whether the response was stored server-side. SambaNova is stateless - always false. nullable: true service_tier: title: Service Tier type: string description: >- The service tier used to process this request, as reported by the server. nullable: true user: title: User type: string description: The user, echoed back from request. nullable: true required: - id - object - status - created_at - model - output - tools - tool_choice - truncation - parallel_tool_calls - background - metadata - temperature - top_p - store - service_tier - error - incomplete_details - frequency_penalty - presence_penalty - user examples: - id: resp_18cb0a9a457241f5b0383ebeee31b80d object: response created_at: 1775853213 completed_at: 1775853214 status: completed model: gpt-oss-120b output: - id: rs_d608e0328e6340a69919f808ee38df6c type: reasoning status: completed summary: [] content: - type: reasoning_text text: >- User wants a three-sentence bedtime story about a unicorn. Provide three sentences, gentle, bedtime. Should be concise. - id: msg_c1eb06dccbc64f469533bb51be664a9e type: message role: assistant status: completed content: - type: output_text annotations: [] logprobs: [] text: >- Under a moonlit canopy of twinkling stars, a silver-mane unicorn tiptoed into the meadow, leaving a trail of soft, glowing stardust behind each gentle step. She whispered a lullaby to the sleepy flowers, and they swayed in rhythm, their petals folding like tiny blankets around the night. As the world hushed, the unicorn curled around a sleepy child's dream, gifting them a night of sweet, sparkling adventures until sunrise. usage: input_tokens: 78 output_tokens: 126 total_tokens: 204 input_tokens_details: cached_tokens: 0 output_tokens_details: reasoning_tokens: 48 error: null incomplete_details: null instructions: null metadata: {} parallel_tool_calls: true tool_choice: auto tools: [] temperature: 1 top_p: 1 frequency_penalty: 0 presence_penalty: 0 reasoning: effort: medium summary: null text: format: type: text truncation: disabled store: false background: false service_tier: free user: null previous_response_id: null - id: resp_e8e3ad78db7546ce969f79b0828f3be3 object: response created_at: 1775856350 completed_at: 1775856350 status: completed model: gpt-oss-120b output: - id: rs_c512ec8087b247b0962b64b7595c92d2 type: reasoning status: completed summary: [] content: - type: reasoning_text text: >- User asks: "What is the weather in Bogotá?" Need to fetch current weather using function get_weather. Provide city name "Bogotá". Use function. - id: fc_c839dacb61a54c10b6560ad8d161f004 type: function_call call_id: call_e52f086a16c94fa796 name: get_weather arguments: '{"city":"Bogotá"}' status: completed usage: input_tokens: 128 output_tokens: 59 total_tokens: 187 input_tokens_details: cached_tokens: 0 output_tokens_details: reasoning_tokens: 43 error: null incomplete_details: null instructions: null metadata: {} parallel_tool_calls: true tool_choice: auto tools: - type: function name: get_weather description: Get the current weather for a city parameters: type: object properties: city: type: string required: - city strict: null temperature: 1 top_p: 1 frequency_penalty: 0 presence_penalty: 0 reasoning: effort: medium summary: null text: format: type: text truncation: disabled store: false background: false service_tier: free user: null previous_response_id: null - id: resp_673393d2de28449581eb6190a8168f84 object: response created_at: 1775517028 completed_at: 1775517028 status: completed model: gpt-oss-120b output: - id: rs_0c34dec41df941b2bce83e11a7614d31 type: reasoning status: completed summary: [] content: - type: reasoning_text text: >- The user wants to extract event info from a sentence. We need to output a JSON in the prescribed format. - id: msg_3b1e13019bd84ebcbcf0c15f45f91dfe type: message role: assistant status: completed content: - type: output_text annotations: [] logprobs: [] text: | { "title": "SambaNova demo", "date": "2026-04-18", "time": "15:30", "location": "San Francisco" } usage: input_tokens: 131 output_tokens: 280 total_tokens: 411 input_tokens_details: cached_tokens: 0 output_tokens_details: reasoning_tokens: 18 error: null incomplete_details: null instructions: null metadata: {} parallel_tool_calls: true tool_choice: auto tools: [] temperature: 1 top_p: 1 frequency_penalty: 0 presence_penalty: 0 reasoning: effort: medium summary: null text: format: type: json_schema name: event_extraction schema: type: object additionalProperties: false properties: title: type: string date: type: string time: type: string location: type: string required: - title - date - time - location truncation: disabled store: false background: false service_tier: free user: null previous_response_id: null ResponseStreamEvent: title: Response Stream Event description: >- Top-level discriminated union of all Server-Sent Events emitted during a streaming `POST /responses` call. Discriminated by the `type` field. Events arrive in `sequence_number` order and cover the full lifecycle of a response: creation, content generation (text, reasoning, function call arguments), and completion. oneOf: - $ref: '#/components/schemas/ResponseCreatedEvent' - $ref: '#/components/schemas/ResponseInProgressEvent' - $ref: '#/components/schemas/ResponseOutputItemAddedEvent' - $ref: '#/components/schemas/ResponseContentPartAddedEvent' - $ref: '#/components/schemas/ResponseReasoningTextDeltaEvent' - $ref: '#/components/schemas/ResponseReasoningTextDoneEvent' - $ref: '#/components/schemas/ResponseOutputTextDeltaEvent' - $ref: '#/components/schemas/ResponseOutputTextDoneEvent' - $ref: '#/components/schemas/ResponseFunctionCallArgumentsDeltaEvent' - $ref: '#/components/schemas/ResponseFunctionCallArgumentsDoneEvent' - $ref: '#/components/schemas/ResponseContentPartDoneEvent' - $ref: '#/components/schemas/ResponseOutputItemDoneEvent' - $ref: '#/components/schemas/ResponseCompletedEvent' discriminator: propertyName: type mapping: response.created: $ref: '#/components/schemas/ResponseCreatedEvent' response.in_progress: $ref: '#/components/schemas/ResponseInProgressEvent' response.output_item.added: $ref: '#/components/schemas/ResponseOutputItemAddedEvent' response.content_part.added: $ref: '#/components/schemas/ResponseContentPartAddedEvent' response.reasoning_text.delta: $ref: '#/components/schemas/ResponseReasoningTextDeltaEvent' response.reasoning_text.done: $ref: '#/components/schemas/ResponseReasoningTextDoneEvent' response.output_text.delta: $ref: '#/components/schemas/ResponseOutputTextDeltaEvent' response.output_text.done: $ref: '#/components/schemas/ResponseOutputTextDoneEvent' response.function_call_arguments.delta: $ref: '#/components/schemas/ResponseFunctionCallArgumentsDeltaEvent' response.function_call_arguments.done: $ref: '#/components/schemas/ResponseFunctionCallArgumentsDoneEvent' response.content_part.done: $ref: '#/components/schemas/ResponseContentPartDoneEvent' response.output_item.done: $ref: '#/components/schemas/ResponseOutputItemDoneEvent' response.completed: $ref: '#/components/schemas/ResponseCompletedEvent' GeneralError: title: GeneralError type: object description: other kind of errors properties: error: type: object properties: code: title: code type: string description: error code nullable: true message: title: message type: string description: error message param: title: param type: string description: error params nullable: true type: title: type type: string description: error type error_model_output: title: Error Model Output type: string description: >- Raw model output that could not be parsed. Present on `server_error` responses when the model produced output that failed internal parsing (e.g. a malformed tool call JSON). nullable: true request_id: title: request_id type: string description: unique request identifier for debugging nullable: true required: - error SimpleError: title: SimpleError type: object description: other kind of simple schema errors properties: error: title: error type: string description: error detail. nullable: true required: - error ResponseInputItem: title: Response Input Item description: >- A single item in the input[] array. Can be a message, a function call, a function call result, or a reasoning trace replayed from a prior turn. When building multi-turn conversations, pass all items from the previous response output[] back as input[] items (including reasoning items) so the model has full context. Discriminated by the "type" field. oneOf: - $ref: '#/components/schemas/ResponseMessage' - $ref: '#/components/schemas/ResponseFunctionCall' - $ref: '#/components/schemas/ResponseFunctionCallOutput' - $ref: '#/components/schemas/ResponseOutputReasoning' discriminator: propertyName: type mapping: message: $ref: '#/components/schemas/ResponseMessage' function_call: $ref: '#/components/schemas/ResponseFunctionCall' function_call_output: $ref: '#/components/schemas/ResponseFunctionCallOutput' reasoning: $ref: '#/components/schemas/ResponseOutputReasoning' ResponseTool: title: Response Tool type: object description: >- A tool definition passed to the model via the tools[] array. Only type "function" is supported; other tool types are filtered server-side. properties: type: title: Type type: string description: The type of tool. Always "function". enum: - function name: title: Name type: string description: The name of the function to be called. description: title: Description type: string description: >- A description of what the function does. Used by the model to decide when and whether to call it. nullable: true parameters: title: Parameters type: object description: >- The parameters the function accepts, described as a JSON Schema object. Omit or set to null for functions that take no parameters. additionalProperties: true nullable: true strict: title: Strict type: boolean description: >- When true, the model is instructed to generate arguments that strictly conform to the provided JSON schema. Not currently enforced by SambaNova accepted for API compatibility and echoed back as null in the response. nullable: true required: - type - name example: type: function name: get_weather description: Get the current weather for a given location. parameters: type: object properties: location: type: string description: The city and state, e.g. San Francisco, CA required: - location additionalProperties: false strict: null ResponseToolChoiceOption: title: Response Tool Choice Option description: >- Controls which tool (if any) the model calls. String values: "none" disables tool calls; "auto" lets the model decide; "required" forces at least one tool call. To force a specific function, provide a ResponseNamedToolChoice object. oneOf: - type: string title: Tool choice mode description: String shorthand for tool selection behavior. enum: - none - auto - required - $ref: '#/components/schemas/ResponseNamedToolChoice' ResponseFormatConfiguration: title: Response Format Configuration description: >- Controls the format of the model's output for the Responses API text.format field. "text" returns plain text (default). "json_object" returns a raw JSON object. "json_schema" enforces structured output matching a provided JSON Schema definition (fields are flat — no json_schema wrapper key). oneOf: - $ref: '#/components/schemas/ResponseTextFormatText' - $ref: '#/components/schemas/ResponseTextFormatJSONObject' - $ref: '#/components/schemas/ResponseTextFormatJSONSchema' discriminator: propertyName: type mapping: text: $ref: '#/components/schemas/ResponseTextFormatText' json_object: $ref: '#/components/schemas/ResponseTextFormatJSONObject' json_schema: $ref: '#/components/schemas/ResponseTextFormatJSONSchema' ResponseOutputItem: title: Response Output Item description: >- A single item in the output[] array returned by the model. Can be a message (with text or tool call content), a function call request, or a reasoning trace. Discriminated by the "type" field. oneOf: - $ref: '#/components/schemas/ResponseMessage' - $ref: '#/components/schemas/ResponseFunctionCall' - $ref: '#/components/schemas/ResponseOutputReasoning' discriminator: propertyName: type mapping: message: $ref: '#/components/schemas/ResponseMessage' function_call: $ref: '#/components/schemas/ResponseFunctionCall' reasoning: $ref: '#/components/schemas/ResponseOutputReasoning' ResponseUsage: title: Response Usage type: object description: Token usage statistics for this response. properties: input_tokens: title: Input Tokens type: integer description: >- Number of tokens in the input, including conversation history and any system instructions. output_tokens: title: Output Tokens type: integer description: >- Number of tokens generated in the output, including visible text and any reasoning tokens. total_tokens: title: Total Tokens type: integer description: Total tokens consumed by this request (input + output). input_tokens_details: title: Input Tokens Details type: object description: Breakdown of input token consumption. properties: cached_tokens: title: Cached Tokens type: integer description: >- Number of input tokens served from the prompt cache. Cached tokens are billed at a reduced rate. nullable: true output_tokens_details: title: Output Tokens Details type: object description: Breakdown of output token consumption. properties: reasoning_tokens: title: Reasoning Tokens type: integer description: >- Number of tokens consumed by the model's internal reasoning process. Only present on reasoning-capable models. nullable: true start_time: title: Start Time type: number description: Unix timestamp (seconds) of when generation started. nullable: true end_time: title: End Time type: number description: Unix timestamp (seconds) of when generation finished. nullable: true time_to_first_token: title: Time To First Token type: number description: Time in seconds from request receipt to first output token (TTFT). nullable: true total_latency: title: Total Latency type: number description: Total time in seconds taken to generate the full response. nullable: true output_tokens_per_sec: title: Output Tokens Per Sec type: number description: Output token throughput (tokens/second) for this response. nullable: true output_tokens_after_first_per_sec: title: Output Tokens After First Per Sec type: number description: >- Output token throughput measured after the first token was emitted. Excludes the time-to-first-token latency from the rate calculation. nullable: true total_tokens_per_sec: title: Total Tokens Per Sec type: number description: Total token throughput (input + output tokens/second). nullable: true acceptance_rate: title: Acceptance Rate type: number description: >- Speculative decoding acceptance rate. Ratio of draft tokens accepted by the verifier model. Higher values indicate better speculation efficiency. nullable: true is_last_response: title: Is Last Response type: boolean description: >- Always true for non-streaming responses. For streaming, true only on the final chunk that carries usage. nullable: true required: - input_tokens - output_tokens - total_tokens example: input_tokens: 248 output_tokens: 72 total_tokens: 320 input_tokens_details: cached_tokens: 0 output_tokens_details: reasoning_tokens: 18 start_time: 1737642515.445 end_time: 1737642515.904 time_to_first_token: 0.084 total_latency: 0.459 output_tokens_per_sec: 156.8 output_tokens_after_first_per_sec: 161.2 total_tokens_per_sec: 311.6 acceptance_rate: 4.06 is_last_response: true ResponseCreatedEvent: title: Response Created Event type: object description: >- Emitted as the first SSE event when a streaming response begins. The embedded response object has status `in_progress` and an empty `output` array. Clients can use this event to capture the response `id` for later correlation. properties: type: type: string enum: - response.created sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. Starts at 0 for `response.created`. response: $ref: '#/components/schemas/ResponseResponse' required: - type - sequence_number - response example: type: response.created sequence_number: 0 response: id: resp_18cb0a9a457241f5b0383ebeee31b80d object: response created_at: 1775853213 status: in_progress model: gpt-oss-120b output: [] usage: null error: null incomplete_details: null instructions: null metadata: {} parallel_tool_calls: true tool_choice: auto tools: [] temperature: 1 top_p: 1 frequency_penalty: 0 presence_penalty: 0 reasoning: effort: medium summary: null text: format: type: text truncation: disabled store: false background: false service_tier: free ResponseInProgressEvent: title: Response In Progress Event type: object description: >- Emitted when the response transitions to the `in_progress` state, immediately after `response.created`. The embedded response object still has an empty `output` array; model generation has not yet produced any content. properties: type: type: string enum: - response.in_progress sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. response: $ref: '#/components/schemas/ResponseResponse' required: - type - sequence_number - response example: type: response.in_progress sequence_number: 1 response: id: resp_18cb0a9a457241f5b0383ebeee31b80d object: response created_at: 1775853213 status: in_progress model: gpt-oss-120b output: [] usage: null error: null incomplete_details: null instructions: null metadata: {} parallel_tool_calls: true tool_choice: auto tools: [] temperature: 1 top_p: 1 frequency_penalty: 0 presence_penalty: 0 reasoning: effort: medium summary: null text: format: type: text truncation: disabled store: false background: false service_tier: free ResponseOutputItemAddedEvent: title: Response Output Item Added Event type: object description: >- Emitted when a new output item (message, function call, or reasoning) is added to the response output array. The item has `status: in_progress` and may have empty content at this point. properties: type: type: string enum: - response.output_item.added sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. output_index: type: integer description: Index of the output item in the response `output` array. item: $ref: '#/components/schemas/ResponseOutputItem' required: - type - sequence_number - output_index - item examples: - type: response.output_item.added sequence_number: 2 output_index: 0 item: id: rs_d608e0328e6340a69919f808ee38df6c type: reasoning status: in_progress summary: [] content: [] - type: response.output_item.added sequence_number: 8 output_index: 1 item: id: msg_c1eb06dccbc64f469533bb51be664a9e type: message role: assistant status: in_progress content: [] - type: response.output_item.added sequence_number: 8 output_index: 1 item: id: fc_c839dacb61a54c10b6560ad8d161f004 type: function_call call_id: call_e52f086a16c94fa796 name: get_weather arguments: '' status: in_progress ResponseContentPartAddedEvent: title: Response Content Part Added Event type: object description: >- Emitted when a new content part is added to an output item's content array. The part's `text` is empty at this point; subsequent `response.output_text.delta` or `response.reasoning_text.delta` events carry the incremental content. properties: type: type: string enum: - response.content_part.added sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the output item to which this content part belongs. output_index: type: integer description: Index of the parent output item in the response `output` array. content_index: type: integer description: Index of this content part within the output item's `content` array. part: $ref: '#/components/schemas/ResponseStreamContentPart' required: - type - sequence_number - item_id - output_index - content_index - part examples: - type: response.content_part.added sequence_number: 3 item_id: rs_d608e0328e6340a69919f808ee38df6c output_index: 0 content_index: 0 part: type: reasoning_text text: '' - type: response.content_part.added sequence_number: 9 item_id: msg_c1eb06dccbc64f469533bb51be664a9e output_index: 1 content_index: 0 part: type: output_text text: '' annotations: [] logprobs: [] ResponseReasoningTextDeltaEvent: title: Response Reasoning Text Delta Event type: object description: >- Emitted for each incremental chunk of reasoning text generated by the model. Accumulate `delta` values in order to reconstruct the full reasoning text. properties: type: type: string enum: - response.reasoning_text.delta sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the reasoning output item this delta belongs to. output_index: type: integer description: Index of the reasoning output item in the response `output` array. delta: type: string description: The incremental chunk of reasoning text. required: - type - sequence_number - item_id - output_index - delta example: type: response.reasoning_text.delta sequence_number: 4 item_id: rs_d608e0328e6340a69919f808ee38df6c output_index: 0 delta: >- User wants a three-sentence bedtime story about a unicorn. Provide three sentences, gentle, bedtime. Should be concise. ResponseReasoningTextDoneEvent: title: Response Reasoning Text Done Event type: object description: >- Emitted when a reasoning content part has been fully generated. The `text` field contains the complete accumulated reasoning text (equivalent to concatenating all preceding `response.reasoning_text.delta` values). Structural mirror of `response.reasoning_text.delta` with `text` instead of `delta`. properties: type: type: string enum: - response.reasoning_text.done x-stainless-const: true sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the reasoning output item this content belongs to. output_index: type: integer description: Index of the reasoning output item in the response `output` array. text: type: string description: The full accumulated reasoning text for this content part. required: - type - sequence_number - item_id - output_index - text example: type: response.reasoning_text.done sequence_number: 5 item_id: rs_d608e0328e6340a69919f808ee38df6c output_index: 0 text: >- User wants a three-sentence bedtime story about a unicorn. Provide three sentences, gentle, bedtime. Should be concise. ResponseOutputTextDeltaEvent: title: Response Output Text Delta Event type: object description: >- Emitted for each incremental chunk of assistant message text. Accumulate `delta` values in order to reconstruct the full output text. properties: type: type: string enum: - response.output_text.delta sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the message output item this delta belongs to. output_index: type: integer description: Index of the message output item in the response `output` array. content_index: type: integer description: Index of the content part within the output item's `content` array. delta: type: string description: The incremental chunk of output text. annotations: type: array items: $ref: '#/components/schemas/ResponseAnnotation' description: Annotations attached to this delta chunk, if any. logprobs: type: array items: $ref: '#/components/schemas/LogProbsContent' description: >- Log probability information for the tokens in this delta, if requested. required: - type - sequence_number - item_id - output_index - content_index - delta - annotations - logprobs example: type: response.output_text.delta sequence_number: 10 item_id: msg_c1eb06dccbc64f469533bb51be664a9e output_index: 1 content_index: 0 delta: >- Under a moonlit canopy of twinkling stars, a silver-mane unicorn tiptoed into the meadow, leaving a trail of soft annotations: [] logprobs: [] ResponseOutputTextDoneEvent: title: Response Output Text Done Event type: object description: >- Emitted when an output text content part has been fully generated. The `text` field contains the complete accumulated output text (equivalent to concatenating all preceding `response.output_text.delta` values). Structural mirror of `response.output_text.delta` with `text` instead of `delta`. properties: type: type: string enum: - response.output_text.done sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the message output item this content belongs to. output_index: type: integer description: Index of the message output item in the response `output` array. content_index: type: integer description: Index of the content part within the output item's `content` array. text: type: string description: The full accumulated output text for this content part. annotations: type: array items: $ref: '#/components/schemas/ResponseAnnotation' description: All annotations attached to this text part. logprobs: type: array items: $ref: '#/components/schemas/LogProbsContent' description: Log probability information for all tokens, if requested. required: - type - sequence_number - item_id - output_index - content_index - text - annotations - logprobs example: type: response.output_text.done sequence_number: 15 item_id: msg_c1eb06dccbc64f469533bb51be664a9e output_index: 1 content_index: 0 annotations: [] logprobs: [] text: >- Under a moonlit canopy of twinkling stars, a silver-mane unicorn tiptoed into the meadow, leaving a trail of soft, glowing stardust behind each gentle step. She whispered a lullaby to the sleepy flowers, and they swayed in rhythm, their petals folding like tiny blankets around the night. As the world hushed, the unicorn curled around a sleepy child's dream, gifting them a night of sweet, sparkling adventures until sunrise. ResponseFunctionCallArgumentsDeltaEvent: title: Response Function Call Arguments Delta Event type: object description: >- Emitted for each incremental chunk of function call arguments JSON generated by the model. Accumulate `delta` values in order to reconstruct the full arguments string. The accumulated result will be a valid JSON string once the corresponding `response.function_call_arguments.done` event is received. properties: type: type: string enum: - response.function_call_arguments.delta sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the function call output item this delta belongs to. output_index: type: integer description: >- Index of the function call output item in the response `output` array. delta: type: string description: The incremental chunk of the function call arguments JSON string. required: - type - sequence_number - item_id - output_index - delta example: type: response.function_call_arguments.delta sequence_number: 9 item_id: fc_c839dacb61a54c10b6560ad8d161f004 output_index: 1 delta: '{"city":"Bogotá"}' ResponseFunctionCallArgumentsDoneEvent: title: Response Function Call Arguments Done Event type: object description: >- Emitted when a function call's arguments have been fully generated. The `arguments` field contains the complete JSON arguments string (equivalent to concatenating all preceding `response.function_call_arguments.delta` values). Structural mirror of `response.function_call_arguments.delta` with `arguments` instead of `delta`. properties: type: type: string enum: - response.function_call_arguments.done sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the function call output item. output_index: type: integer description: >- Index of the function call output item in the response `output` array. arguments: type: string description: The complete JSON arguments string for the function call. required: - type - sequence_number - item_id - output_index - arguments example: type: response.function_call_arguments.done sequence_number: 10 item_id: fc_c839dacb61a54c10b6560ad8d161f004 output_index: 1 arguments: '{"city":"Bogotá"}' ResponseContentPartDoneEvent: title: Response Content Part Done Event type: object description: >- Emitted when a content part has been fully generated. The `part` field contains the complete content for this part (equivalent to the state after accumulating all preceding delta events). Structural mirror of `response.content_part.added`. properties: type: type: string enum: - response.content_part.done sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. item_id: type: string description: ID of the output item this content part belongs to. output_index: type: integer description: Index of the parent output item in the response `output` array. content_index: type: integer description: Index of this content part within the output item's `content` array. part: $ref: '#/components/schemas/ResponseStreamContentPart' required: - type - sequence_number - item_id - output_index - content_index - part examples: - type: response.content_part.done sequence_number: 6 item_id: rs_d608e0328e6340a69919f808ee38df6c output_index: 0 content_index: 0 part: type: reasoning_text text: >- User wants a three-sentence bedtime story about a unicorn. Provide three sentences, gentle, bedtime. Should be concise. - type: response.content_part.done sequence_number: 16 item_id: msg_c1eb06dccbc64f469533bb51be664a9e output_index: 1 content_index: 0 part: type: output_text annotations: [] logprobs: [] text: >- Under a moonlit canopy of twinkling stars, a silver-mane unicorn tiptoed into the meadow, leaving a trail of soft, glowing stardust behind each gentle step. She whispered a lullaby to the sleepy flowers, and they swayed in rhythm, their petals folding like tiny blankets around the night. As the world hushed, the unicorn curled around a sleepy child's dream, gifting them a night of sweet, sparkling adventures until sunrise. ResponseOutputItemDoneEvent: title: Response Output Item Done Event type: object description: >- Emitted when an output item has been fully generated and its status transitions to `completed`. The `item` field contains the complete output item including all accumulated content. Structural mirror of `response.output_item.added`. properties: type: type: string enum: - response.output_item.done sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. output_index: type: integer description: Index of the output item in the response `output` array. item: $ref: '#/components/schemas/ResponseOutputItem' required: - type - sequence_number - output_index - item examples: - type: response.output_item.done sequence_number: 7 output_index: 0 item: id: rs_d608e0328e6340a69919f808ee38df6c type: reasoning status: completed summary: [] content: - type: reasoning_text text: >- User wants a three-sentence bedtime story about a unicorn. Provide three sentences, gentle, bedtime. Should be concise. - type: response.output_item.done sequence_number: 17 output_index: 1 item: id: msg_c1eb06dccbc64f469533bb51be664a9e type: message role: assistant status: completed content: - type: output_text annotations: [] logprobs: [] text: >- Under a moonlit canopy of twinkling stars, a silver-mane unicorn tiptoed into the meadow, leaving a trail of soft, glowing stardust behind each gentle step. She whispered a lullaby to the sleepy flowers, and they swayed in rhythm, their petals folding like tiny blankets around the night. As the world hushed, the unicorn curled around a sleepy child's dream, gifting them a night of sweet, sparkling adventures until sunrise. - type: response.output_item.done sequence_number: 11 output_index: 1 item: id: fc_c839dacb61a54c10b6560ad8d161f004 type: function_call call_id: call_e52f086a16c94fa796 name: get_weather arguments: '{"city":"Bogotá"}' status: completed ResponseCompletedEvent: title: Response Completed Event type: object description: >- Emitted as the final SSE event when a streaming response has been fully generated. The embedded response object has `status: completed`, a fully populated `output` array, and a populated `usage` object. Structural mirror of `response.created`. properties: type: type: string enum: - response.completed x-stainless-const: true sequence_number: type: integer description: >- Monotonically increasing counter for ordering events within a streaming response. response: $ref: '#/components/schemas/ResponseResponse' required: - type - sequence_number - response examples: - type: response.completed sequence_number: 18 response: id: resp_18cb0a9a457241f5b0383ebeee31b80d object: response created_at: 1775853213 completed_at: 1775853214 status: completed model: gpt-oss-120b output: - id: rs_d608e0328e6340a69919f808ee38df6c type: reasoning status: completed summary: [] content: - type: reasoning_text text: >- User wants a three-sentence bedtime story about a unicorn. Provide three sentences, gentle, bedtime. Should be concise. - id: msg_c1eb06dccbc64f469533bb51be664a9e type: message role: assistant status: completed content: - type: output_text annotations: [] logprobs: [] text: >- Under a moonlit canopy of twinkling stars, a silver-mane unicorn tiptoed into the meadow, leaving a trail of soft, glowing stardust behind each gentle step. usage: input_tokens: 78 output_tokens: 126 total_tokens: 204 input_tokens_details: cached_tokens: 0 tool_choice: auto tools: [] temperature: 1 top_p: 1 reasoning: effort: medium summary: null text: format: type: text - type: response.completed sequence_number: 12 response: id: resp_e8e3ad78db7546ce969f79b0828f3be3 object: response created_at: 1775856350 completed_at: 1775856350 status: completed model: gpt-oss-120b output: - id: rs_c512ec8087b247b0962b64b7595c92d2 type: reasoning status: completed summary: [] content: - type: reasoning_text text: >- User asks: "What is the weather in Bogotá?" Need to fetch current weather using function get_weather. - id: fc_c839dacb61a54c10b6560ad8d161f004 type: function_call call_id: call_e52f086a16c94fa796 name: get_weather arguments: '{"city":"Bogotá"}' status: completed usage: input_tokens: 128 output_tokens: 59 total_tokens: 187 input_tokens_details: cached_tokens: 0 output_tokens_details: reasoning_tokens: 43 tools: - type: function name: get_weather description: Get the current weather for a city parameters: type: object properties: city: type: string required: - city strict: null - type: response.completed sequence_number: 14 response: id: resp_673393d2de28449581eb6190a8168f84 object: response created_at: 1775517028 completed_at: 1775517028 status: completed model: gpt-oss-120b output: - id: rs_0c34dec41df941b2bce83e11a7614d31 type: reasoning status: completed summary: [] content: - type: reasoning_text text: >- The user wants to extract event info from a sentence. We need to output a JSON in the prescribed format. - id: msg_3b1e13019bd84ebcbcf0c15f45f91dfe type: message role: assistant status: completed content: - type: output_text annotations: [] logprobs: [] text: | { "title": "SambaNova demo", "date": "2026-04-18", "time": "15:30", "location": "San Francisco" } usage: input_tokens: 131 output_tokens: 280 total_tokens: 411 input_tokens_details: cached_tokens: 0 text: format: type: json_schema name: event_extraction schema: type: object additionalProperties: false properties: title: type: string date: type: string time: type: string location: type: string required: - title - date - time - location ResponseMessage: title: Response Message type: object description: >- A message item. When used as input, id and status are optional. When present in output[], id and status are always set by the server. Role "assistant" with content type "output_text" represents a prior model turn; user/system/developer turns use content type "input_text". Plain string content is accepted in all roles on input. properties: type: title: Type type: string description: The type of this item. Always "message". enum: - message nullable: true id: title: ID type: string description: >- Unique identifier for this message item. Always present on output items returned by the server. Not required when constructing input items. nullable: true role: title: Role type: string description: >- The role of the message. "assistant" is set by the server on output items. All four values are valid when constructing input items. enum: - user - assistant - system - developer content: title: Content description: >- The message content. Either a plain text string or an array of typed content parts. For assistant turns, array parts use type "output_text"; for user/system/developer turns, array parts use type "input_text". oneOf: - type: string title: Text content description: A plain text message body. - type: array title: Content part array description: An array of typed content parts. items: $ref: '#/components/schemas/ResponseContent' status: title: Status type: string description: >- Processing state of this item. Always set on output items returned by the server. Not required when constructing input items. enum: - in_progress - completed - incomplete required: - type - role - content examples: - type: message role: user content: What is the weather in San Francisco? - type: message role: user content: - type: input_text text: Describe what you see in this image. - type: message role: assistant content: - type: output_text text: The weather in San Francisco is 65°F and partly cloudy. - type: message role: system content: You are a helpful assistant that provides weather information. - type: message role: developer content: Always respond in JSON format when asked for structured data. ResponseFunctionCall: title: Response Function Call type: object description: >- A function call item used in both input[] and output[]. When present in output[], the model is requesting a tool call — id and status are always set by the server. When used in input[], it replays a prior function call from the conversation history for multi-turn tool loops. properties: type: title: Type type: string description: The type of this item. Always "function_call". enum: - function_call id: title: ID type: string description: >- Unique identifier for this function call item. Always present on output items returned by the server. Required when replaying this item in a subsequent input[]. nullable: true call_id: title: Call ID type: string description: >- The call ID generated by the model for this tool invocation. Used to correlate the function call with its corresponding function_call_output result. name: title: Name type: string description: The name of the function the model wants to call. arguments: title: Arguments type: string description: >- A JSON string of the arguments to pass to the function, as generated by the model. Validate the arguments in your code before calling the function — the model may hallucinate parameters not defined in the schema. status: title: Status type: string description: >- Processing state of this item. Always set on output items returned by the server. Not required when constructing input items. enum: - in_progress - completed - incomplete nullable: true required: - type - call_id - name - arguments example: type: function_call id: fc_abc123 call_id: call_abc123 name: get_weather arguments: '{"location":"San Francisco, CA"}' status: completed ResponseFunctionCallOutput: title: Response Function Call Output type: object description: >- A tool result item used in input[] to return the output of a function call back to the model. Always input-only — the model never outputs function_call_output items. The call_id must match the id of the corresponding ResponseFunctionCall item from the prior model turn. properties: type: title: Type type: string enum: - function_call_output description: The type of this item. Always "function_call_output". id: title: ID type: string description: Optional identifier for this output item. nullable: true call_id: title: Call ID type: string description: >- The call_id of the ResponseFunctionCall item this result corresponds to. Must match exactly. output: title: Output type: string description: >- A JSON string containing the result of the function call. Returned to the model as the tool's response. status: title: Status type: string enum: - in_progress - completed - incomplete description: Processing state. Optional when constructing input items. nullable: true required: - type - call_id - output example: type: function_call_output call_id: call_abc123 output: '{"temperature":"65°F","condition":"partly cloudy"}' ResponseOutputReasoning: title: Response Output Reasoning type: object description: >- A reasoning item emitted in output[] by models that expose their internal chain-of-thought. Contains the raw reasoning text the model produced before generating its final answer. properties: type: title: Type type: string description: The type of this output item. Always "reasoning". enum: - reasoning id: title: ID type: string description: Unique identifier for this reasoning item, assigned by the server. nullable: true status: title: Status type: string description: >- Processing state of this item. "in_progress" while streaming; "completed" when the full reasoning text has been emitted. enum: - in_progress - completed - incomplete nullable: true content: title: Content type: array description: >- The reasoning content parts. Each part has type "reasoning_text" and carries the raw thinking text. Absent when the item is first added (`response.output_item.added`); populated by `response.output_item.done`. items: $ref: '#/components/schemas/ResponseReasoningContent' nullable: true summary: title: Summary type: array description: >- A condensed summary of the reasoning content. Always returned as an empty array. Included for OpenAI API compatibility. items: type: object additionalProperties: true nullable: true required: - type example: type: reasoning id: rs_msg_5dc0dc6fca844a20885869634cd3014a status: completed summary: [] content: - type: reasoning_text text: >- The user wants weather information for San Francisco. I'll call get_weather with location "San Francisco, CA". ResponseNamedToolChoice: title: Response Named Tool Choice type: object description: >- Forces the model to call a specific function. Pass this object as tool_choice to guarantee the model will invoke the named function. properties: type: title: Type type: string description: The type of tool to force. Always "function". enum: - function name: title: Name type: string description: The name of the function to call. required: - type - name example: type: function name: get_weather ResponseTextFormatText: title: Response Text Format Text type: object description: >- Specifies that the model should produce output as plain text. This is the default. additionalProperties: false properties: type: title: Type type: string description: The format type. Always "text". enum: - text required: - type example: type: text ResponseTextFormatJSONObject: title: Response Text Format JSON Object type: object description: >- Specifies that the model should produce output as a raw JSON object. The model is not constrained to any specific schema. additionalProperties: false properties: type: title: Type type: string description: The format type. Always "json_object". enum: - json_object required: - type example: type: json_object ResponseTextFormatJSONSchema: title: Response Text Format JSON Schema type: object description: >- Specifies that the model should produce structured output conforming to the provided JSON Schema. Unlike the Chat Completions API, the Responses API places name, schema, strict, and description directly on this object (not nested under a json_schema key). additionalProperties: false properties: type: title: Type type: string description: The format type. Always "json_schema". enum: - json_schema name: title: Name type: string description: A name for the schema, used to identify it in the model's output. description: title: Description type: string description: An optional description of the schema, used to guide the model. nullable: true schema: title: Schema type: object description: >- A JSON Schema object defining the structure the model's output must conform to. Typically includes "type", "properties", "required", and optionally "additionalProperties". Must be a valid JSON Schema. additionalProperties: true nullable: true strict: title: Strict type: boolean description: >- When true, the model is instructed to produce output that strictly matches the schema. Not currently enforced by SambaNova — accepted for API compatibility. nullable: true required: - type - name example: type: json_schema name: event_extraction description: Extract structured event information from text strict: false schema: type: object properties: title: type: string date: type: string time: type: string location: type: string required: - title - date - time - location additionalProperties: false ResponseStreamContentPart: title: Response Stream Content Part description: >- A content part within a streaming event. Discriminated by `type`: `output_text` for assistant message content (reuses ResponseTextContent), `reasoning_text` for model reasoning (OSS extension). oneOf: - $ref: '#/components/schemas/ResponseTextContent' - $ref: '#/components/schemas/ResponseReasoningContent' discriminator: propertyName: type mapping: output_text: $ref: '#/components/schemas/ResponseTextContent' reasoning_text: $ref: '#/components/schemas/ResponseReasoningContent' ResponseAnnotation: title: Response Annotation description: >- An annotation on a span of output text. Currently stubbed for OpenAI API compatibility — SambaNova does not produce annotations. oneOf: - $ref: '#/components/schemas/ResponseUrlCitation' - $ref: '#/components/schemas/ResponseFileCitation' discriminator: propertyName: type mapping: url_citation: $ref: '#/components/schemas/ResponseUrlCitation' file_citation: $ref: '#/components/schemas/ResponseFileCitation' LogProbsContent: title: Log Probs Content additionalProperties: true properties: bytes: title: Bytes anyOf: - items: type: integer type: array nullable: true logprob: title: Logprob type: number token: title: Token type: string top_logprobs: $ref: '#/components/schemas/TopLogProbs' required: - token - logprob - top_logprobs examples: - token: ' Hello' logprob: -0.00012340000000000002 bytes: - 32 - 72 - 101 - 108 - 108 - 111 top_logprobs: - token: ' Hello' logprob: -0.00012340000000000002 bytes: - 32 - 72 - 101 - 108 - 108 - 111 - token: ' Hi' logprob: -8.243 bytes: - 32 - 72 - 105 ResponseContent: title: Response Content description: >- A typed content part used in both input and output message content arrays. Discriminated by the "type" field. oneOf: - $ref: '#/components/schemas/ResponseTextContent' - $ref: '#/components/schemas/ResponseImageContent' - $ref: '#/components/schemas/ResponseAudioContent' - $ref: '#/components/schemas/ResponseVideoContent' discriminator: propertyName: type mapping: input_text: $ref: '#/components/schemas/ResponseTextContent' output_text: $ref: '#/components/schemas/ResponseTextContent' input_image: $ref: '#/components/schemas/ResponseImageContent' input_audio: $ref: '#/components/schemas/ResponseAudioContent' input_video: $ref: '#/components/schemas/ResponseVideoContent' ResponseReasoningContent: title: Response Reasoning Content type: object description: >- A single content part within a reasoning item's content[]. Contains the raw reasoning text produced by the model during its internal thinking process. properties: type: title: Type type: string description: The type of this reasoning content part. Always "reasoning_text". enum: - reasoning_text text: title: Text type: string description: The raw reasoning text generated by the model. required: - type - text example: type: reasoning_text text: >- The user is asking about the weather. I should call the get_weather function with the location extracted from the message. ResponseTextContent: title: Response Input Text Content type: object description: A plain text content part for use inside a ResponseMessage. properties: type: title: Type type: string description: >- The type of this content part. "input_text" for user, developer, and system roles, and output_text for assistant messages. enum: - input_text - output_text text: title: Text type: string description: The text content of this part. annotations: title: Annotations type: array description: >- Annotations on spans of this text part, such as URL or file citations. Only present on output_text parts returned by the server. Not currently produced by SambaNova — included for OpenAI API compatibility. items: $ref: '#/components/schemas/ResponseAnnotation' nullable: true logprobs: title: Log Probs type: array description: >- Log probability information for the output tokens. Only present on output_text parts when `top_logprobs` is set on the request. items: $ref: '#/components/schemas/LogProbsContent' nullable: true required: - type - text ResponseUrlCitation: title: Response URL Citation type: object description: >- A URL citation annotation on an output text span. Not currently implemented by SambaNova — included for OpenAI API compatibility. properties: type: title: Type type: string description: The annotation type. Always "url_citation". enum: - url_citation url: title: URL type: string description: The URL being cited. title: title: Title type: string description: The title of the cited page. nullable: true start_index: title: Start Index type: integer description: Character offset in the output text where the citation starts. end_index: title: End Index type: integer description: Character offset in the output text where the citation ends. required: - type - url - start_index - end_index ResponseFileCitation: title: Response File Citation type: object description: >- A file citation annotation on an output text span. Not currently implemented by SambaNova — included for OpenAI API compatibility. properties: type: title: Type type: string description: The annotation type. Always "file_citation". enum: - file_citation file_id: title: File ID type: string description: The ID of the cited file. start_index: title: Start Index type: integer description: Character offset in the output text where the citation starts. end_index: title: End Index type: integer description: Character offset in the output text where the citation ends. required: - type - file_id - start_index - end_index TopLogProbs: title: TopLogProbs type: object additionalProperties: true properties: bytes: title: Bytes anyOf: - items: type: integer type: array - type: 'null' logprob: title: Logprob type: number token: title: Token type: string required: - token - logprob examples: - token: ' Hello' logprob: -0.00012340000000000002 bytes: - 32 - 72 - 101 - 108 - 108 - 111 - token: ' Hi' logprob: -8.243 bytes: - 32 - 72 - 105 ResponseImageContent: title: Response Image Content type: object description: >- An image content part for use inside a ResponseMessage. Supported in input messages (user role). Provide a base64-encoded image as a data URL. external URLs not supported. properties: type: title: Type type: string description: The type of this content part. Always "input_image". enum: - input_image image_url: title: Image URL type: string description: >- A base64-encoded image in data URL format or fully qualified URL (e.g. "data:image/png;base64,..."). Currently only base64-encoded image data is supported. nullable: true detail: title: Detail type: string description: >- The detail level for image processing. "auto" lets the model decide. "low" uses a fixed low-res tile. "high" enables high-res tiling. enum: - auto - low - high default: auto nullable: true required: - type - image_url ResponseAudioContent: title: Response Audio Content type: object description: >- An audio content part for use inside a ResponseMessage. Supported in input messages (user role). Provide base64-encoded audio data with the format. properties: type: title: Type type: string description: The type of this content part. Always "input_audio". enum: - input_audio input_audio: title: Input Audio type: object description: The audio data and format. properties: data: title: Data type: string description: Base64-encoded audio data. format: title: Format type: string enum: - mp3 - wav description: The format of the audio data. required: - data - format required: - type - input_audio ResponseVideoContent: title: Response Video Content type: object description: >- A video content part for use inside a ResponseMessage. Supported in input messages (user role). Provide a URL or base64-encoded video data. properties: type: title: Type type: string description: The type of this content part. Always "input_video". enum: - input_video video_url: title: Video URL type: string description: >- A URL of the video or base64-encoded video data in data URL format (e.g. "data:video/mp4;base64,..."). required: - type - video_url securitySchemes: api_key: type: http description: >- SambaNova API key, sent as a bearer token in the `Authorization` header (`Authorization: Bearer `). Default authentication scheme used by the SambaNova SDK across every OpenAI compatible endpoint. scheme: bearer bearerFormat: apiKey ````