> ## Documentation Index > Fetch the complete documentation index at: https://docs.linkmodel.ai/llms.txt > Use this file to discover all available pages before exploring further. # OpenAI Decisions > Evaluate shared text, images, or JSON with GPT-6 Luna. Ask predicate, choice, or score questions and receive typed answers in question order. Any question can return a refusal while other questions still receive answers. Use Responses for generated text, custom JSON output schemas, or tool calling. Decisions returns a JSON result; it does not expose a streaming parameter. ### Billing OpenAI's Decisions pricing uses the standard GPT-6 Luna input and cache rates. Output tokens are not charged for Decisions; this does not change Luna's output rate on Chat Completions or Responses. Regional processing and long-context input multipliers still apply. See [LinkModel Chat pricing](/pricing/chat) for the current model rates. Cached and cache-write tokens are included in usage.input_tokens. Subtract those categories before applying the uncached input rate; do not charge them a second time. The example usage below is illustrative, not a token estimate. Official references: [Decisions guide](https://developers.openai.com/api/docs/guides/decisions) and [Create a decision](https://developers.openai.com/api/reference/resources/decisions/methods/create). ## OpenAPI ````yaml /openapi/chat/_openai-decisions.yaml post /decisions openapi: 3.0.3 info: title: OpenAI Decisions version: 1.0.0 description: >- Classification and scoring with GPT-6 Luna through the OpenAI-compatible Decisions endpoint. servers: - url: https://api.linkmodel.ai/v1 description: Production security: - bearerAuth: [] tags: - name: chat paths: /decisions: post: tags: - chat summary: OpenAI Decisions description: > Evaluate shared text, images, or JSON with GPT-6 Luna. Ask predicate, choice, or score questions and receive typed answers in question order. Any question can return a refusal while other questions still receive answers. Use Responses for generated text, custom JSON output schemas, or tool calling. Decisions returns a JSON result; it does not expose a streaming parameter. ### Billing OpenAI's Decisions pricing uses the standard GPT-6 Luna input and cache rates. Output tokens are not charged for Decisions; this does not change Luna's output rate on Chat Completions or Responses. Regional processing and long-context input multipliers still apply. See [LinkModel Chat pricing](/pricing/chat) for the current model rates. Cached and cache-write tokens are included in usage.input_tokens. Subtract those categories before applying the uncached input rate; do not charge them a second time. The example usage below is illustrative, not a token estimate. Official references: [Decisions guide](https://developers.openai.com/api/docs/guides/decisions) and [Create a decision](https://developers.openai.com/api/reference/resources/decisions/methods/create). operationId: openai-decisions requestBody: required: true content: application/json: schema: type: object required: - model - input - questions properties: model: type: string enum: - gpt-6-luna example: gpt-6-luna description: >- Currently only gpt-6-luna supports Decisions. See [Models](/models). input: oneOf: - type: string - type: array items: oneOf: - $ref: '#/components/schemas/UserMessage' - $ref: '#/components/schemas/InstructionMessage' - $ref: '#/components/schemas/AssistantMessage' - $ref: '#/components/schemas/FunctionCall' - $ref: '#/components/schemas/CustomToolCall' - $ref: '#/components/schemas/ToolOutput' description: >- Shared evidence for all questions. Tool outputs must match a preceding call by call_id. Files, audio, and item references are not supported. questions: type: array items: oneOf: - $ref: '#/components/schemas/PredicateQuestion' - $ref: '#/components/schemas/ChoiceQuestion' - $ref: '#/components/schemas/ScoreQuestion' minItems: 1 maxItems: 400 prompt_cache_key: type: string maxLength: 64 nullable: true prompt_cache_options: type: object properties: mode: type: string enum: - implicit - explicit default: implicit ttl: type: string enum: - 30m default: 30m description: >- Cache controls. Explicit mode disables the implicit breakpoint; without explicit breakpoints, no cache is used. Up to four breakpoints can be written per request. safety_identifier: type: string maxLength: 128 nullable: true example: model: gpt-6-luna input: I was charged twice for my order. questions: - type: choice name: department instructions: Which department should handle this request? choices: - value: billing description: Payments, invoices, and refunds. - value: support description: Other product or account questions. responses: '200': description: >- Answers in question order, including per-question refusals, and token usage. content: application/json: schema: type: object required: - answers - model - usage properties: answers: type: array items: oneOf: - $ref: '#/components/schemas/PredicateAnswer' - $ref: '#/components/schemas/ChoiceAnswer' - $ref: '#/components/schemas/ScoreAnswer' - $ref: '#/components/schemas/RefusalAnswer' model: type: string usage: $ref: '#/components/schemas/Usage' example: model: gpt-6-luna answers: - type: choice name: department choice: billing confidence: 0.98 probabilities: - value: billing probability: 0.98 - value: support probability: 0.02 usage: input_tokens: 60 input_tokens_details: cached_tokens: 0 cache_write_tokens: 0 output_tokens: 0 output_tokens_details: reasoning_tokens: 0 total_tokens: 60 '400': description: Bad request -- invalid or missing parameter. content: application/json: schema: type: object properties: code: type: integer example: 400 msg: type: string example: model is required request_id: type: string '401': description: Unauthorized -- missing or invalid API key. content: application/json: schema: type: object properties: code: type: integer example: 401 msg: type: string example: unauthorized request_id: type: string '500': description: Internal server error. content: application/json: schema: type: object properties: code: type: integer example: 500 msg: type: string request_id: type: string components: schemas: UserMessage: type: object required: - role - content properties: type: type: string enum: - message role: type: string enum: - user content: oneOf: - type: string - type: array items: oneOf: - $ref: '#/components/schemas/InputText' - $ref: '#/components/schemas/InputImage' - $ref: '#/components/schemas/InputJSON' InstructionMessage: type: object required: - role - content properties: type: type: string enum: - message role: type: string enum: - developer - system content: oneOf: - type: string - type: array items: $ref: '#/components/schemas/InputText' AssistantMessage: type: object required: - role - content properties: type: type: string enum: - message role: type: string enum: - assistant content: oneOf: - type: string - type: array items: oneOf: - type: object required: - type - text properties: type: type: string enum: - output_text text: type: string - type: object required: - type - refusal properties: type: type: string enum: - refusal refusal: type: string phase: type: string enum: - commentary - final_answer FunctionCall: type: object required: - type - call_id - name - arguments properties: type: type: string enum: - function_call call_id: type: string name: type: string arguments: type: string id: type: string namespace: type: string status: type: string enum: - in_progress - completed - incomplete CustomToolCall: type: object required: - type - call_id - name - input properties: type: type: string enum: - custom_tool_call call_id: type: string name: type: string input: type: string id: type: string namespace: type: string status: type: string enum: - in_progress - completed - incomplete ToolOutput: type: object required: - type - call_id - output properties: type: type: string enum: - function_call_output - custom_tool_call_output call_id: type: string output: oneOf: - type: string - type: array items: oneOf: - $ref: '#/components/schemas/InputText' - $ref: '#/components/schemas/InputImage' id: type: string status: type: string enum: - in_progress - completed - incomplete PredicateQuestion: title: Predicate type: object required: - type - instructions properties: type: type: string enum: - predicate name: type: string description: Optional name echoed in the corresponding answer. Use unique names. instructions: oneOf: - type: string - type: array items: oneOf: - $ref: '#/components/schemas/QuestionText' - $ref: '#/components/schemas/QuestionImage' - $ref: '#/components/schemas/InputJSON' minItems: 1 description: Question instructions. Cache breakpoints are not supported here. ChoiceQuestion: title: Choice type: object required: - type - instructions - choices properties: type: type: string enum: - choice name: type: string description: Optional name echoed in the corresponding answer. Use unique names. instructions: oneOf: - type: string - type: array items: oneOf: - $ref: '#/components/schemas/QuestionText' - $ref: '#/components/schemas/QuestionImage' - $ref: '#/components/schemas/InputJSON' minItems: 1 description: Question instructions. Cache breakpoints are not supported here. choices: type: array items: type: object required: - value properties: value: oneOf: - type: string - type: boolean description: type: string minItems: 2 maxItems: 255 description: >- Unique typed values. A string and a boolean with the same text are distinct. ScoreQuestion: title: Score type: object required: - type - instructions - levels properties: type: type: string enum: - score name: type: string description: Optional name echoed in the corresponding answer. Use unique names. instructions: oneOf: - type: string - type: array items: oneOf: - $ref: '#/components/schemas/QuestionText' - $ref: '#/components/schemas/QuestionImage' - $ref: '#/components/schemas/InputJSON' minItems: 1 description: Question instructions. Cache breakpoints are not supported here. levels: type: array items: type: object required: - label properties: label: type: string description: type: string minItems: 2 maxItems: 10 description: Ordered levels, from lowest to highest. PredicateAnswer: title: Predicate type: object required: - type - name - probability properties: type: type: string enum: - predicate name: type: string nullable: true probability: type: number minimum: 0 maximum: 1 ChoiceAnswer: title: Choice type: object required: - type - name - choice - confidence - probabilities properties: type: type: string enum: - choice name: type: string nullable: true choice: oneOf: - type: string - type: boolean confidence: type: number minimum: 0 maximum: 1 probabilities: type: array items: type: object required: - value - probability properties: value: oneOf: - type: string - type: boolean probability: type: number minimum: 0 maximum: 1 ScoreAnswer: title: Score type: object required: - type - name - score - confidence - probabilities properties: type: type: string enum: - score name: type: string nullable: true score: type: number description: >- Probability-weighted average of the zero-based level indices; may fall between levels. confidence: type: number minimum: 0 maximum: 1 probabilities: type: array items: type: object required: - label - value - probability properties: label: type: string value: type: integer probability: type: number minimum: 0 maximum: 1 RefusalAnswer: title: Refusal type: object required: - type - name properties: type: type: string enum: - refusal name: type: string nullable: true Usage: type: object required: - input_tokens - input_tokens_details - output_tokens - output_tokens_details - total_tokens properties: input_tokens: type: integer minimum: 0 input_tokens_details: type: object required: - cached_tokens - cache_write_tokens properties: cached_tokens: type: integer minimum: 0 cache_write_tokens: type: integer minimum: 0 output_tokens: type: integer minimum: 0 output_tokens_details: type: object required: - reasoning_tokens properties: reasoning_tokens: type: integer minimum: 0 total_tokens: type: integer minimum: 0 InputText: type: object required: - type - text properties: type: type: string enum: - input_text text: type: string prompt_cache_breakpoint: $ref: '#/components/schemas/CacheBreakpoint' InputImage: type: object required: - type - image_url properties: type: type: string enum: - input_image image_url: type: string pattern: ^(data:|https?://) description: >- Public HTTP(S) image URL or base64 data URL. File IDs are not supported. detail: type: string enum: - low - high - auto - original prompt_cache_breakpoint: $ref: '#/components/schemas/CacheBreakpoint' InputJSON: type: object required: - type - data properties: type: type: string enum: - input_json data: oneOf: - type: object additionalProperties: true - type: array items: type: object additionalProperties: true description: >- JSON object or array of objects. At most 32 container levels, 10,000 values, and 256 KiB of compact UTF-8 JSON. QuestionText: type: object required: - type - text properties: type: type: string enum: - input_text text: type: string QuestionImage: type: object required: - type - image_url properties: type: type: string enum: - input_image image_url: type: string pattern: ^(data:|https?://) description: >- Public HTTP(S) image URL or base64 data URL. File IDs are not supported. detail: type: string enum: - low - high - auto - original CacheBreakpoint: type: object required: - mode properties: mode: type: string enum: - explicit securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: API key ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.