> ## Documentation Index > Fetch the complete documentation index at: https://docs.gumloop.com/llms.txt > Use this file to discover all available pages before exploring further. # Create decision > Ask a decision model one or more typed questions about a piece of `state`. The model returns a probability for each question instead of text. The request and response use the OpenRouter Decisions schema. Jev (`typesafe/jev-1.13`) is the first decision model. Any model with `is_classifier_model: true` in `GET /models` works here. Chat models are refused with `model_unavailable`. Send those to `POST /chat/completions`. You choose the question ids. The answers come back under the same ids. There are three question types: - `noul` is a yes/no question. The answer is `noul`, the probability of yes. `criteria.true` and `criteria.false` are optional descriptions of each side. - `choice` picks one option from `criteria`, a map of option id to description. A description may be `null`. The answer is `choice`, `probabilities` over every option, and `confidence`. - `score` places the state on the ordered rubric in `criteria`, a list with the lowest level first. The answer is `score` (the weighted position), `probabilities` per level, `legend`, and `confidence`. `state`, `instructions` and criteria accept a string, a JSON object, or an array. The response `model` is the exact version the provider served, for example `typesafe/jev-1.13-20260917`. Billing uses the model you requested. There is no streaming form. The server applies its zero-data-retention provider policy after any `provider` preferences you send. ### Billing Each call charges the caller's credit balance from the provider-reported cost, or from the model's token rate when the provider reports none. A call costs at least one credit, so put every independent question about one piece of state in a single request. ## OpenAPI ````yaml post /decisions openapi: 3.0.0 info: title: Public API version: 1.0.0 servers: - url: https://api.gumloop.com/api/v1 security: [] paths: /decisions: post: tags: - Models summary: Create decision description: > Ask a decision model one or more typed questions about a piece of `state`. The model returns a probability for each question instead of text. The request and response use the OpenRouter Decisions schema. Jev (`typesafe/jev-1.13`) is the first decision model. Any model with `is_classifier_model: true` in `GET /models` works here. Chat models are refused with `model_unavailable`. Send those to `POST /chat/completions`. You choose the question ids. The answers come back under the same ids. There are three question types: - `noul` is a yes/no question. The answer is `noul`, the probability of yes. `criteria.true` and `criteria.false` are optional descriptions of each side. - `choice` picks one option from `criteria`, a map of option id to description. A description may be `null`. The answer is `choice`, `probabilities` over every option, and `confidence`. - `score` places the state on the ordered rubric in `criteria`, a list with the lowest level first. The answer is `score` (the weighted position), `probabilities` per level, `legend`, and `confidence`. `state`, `instructions` and criteria accept a string, a JSON object, or an array. The response `model` is the exact version the provider served, for example `typesafe/jev-1.13-20260917`. Billing uses the model you requested. There is no streaming form. The server applies its zero-data-retention provider policy after any `provider` preferences you send. ### Billing Each call charges the caller's credit balance from the provider-reported cost, or from the model's token rate when the provider reports none. A call costs at least one credit, so put every independent question about one piece of state in a single request. operationId: createDecision parameters: - in: query name: team_id required: false schema: type: string description: >- Scope the request to a team. When omitted, uses the authenticated user's default organization. requestBody: required: true content: application/json: schema: type: object required: - model - state - questions properties: model: type: string description: >- Decision model id. Use an `id` from `GET /models` whose `is_classifier_model` is `true`. example: typesafe/jev-1.13 state: description: >- The content to evaluate. A plain string, or a JSON object or array of related context. oneOf: - type: string - type: object - type: array items: {} questions: type: object description: >- Questions keyed by ids you choose. Answers come back under the same ids. additionalProperties: oneOf: - type: object title: Yes/no question required: - type - instructions properties: type: type: string enum: - noul instructions: description: The question. A string, or a JSON object or array. criteria: type: object description: What a yes and a no mean. required: - 'true' - 'false' properties: 'true': {} 'false': {} - type: object title: Choice question required: - type - instructions - criteria properties: type: type: string enum: - choice instructions: {} criteria: type: object description: >- Option id to description. A description may be `null`. additionalProperties: true - type: object title: Score question required: - type - instructions - criteria properties: type: type: string enum: - score instructions: {} criteria: type: array description: Ordered rubric, lowest level first. items: {} provider: type: object description: >- OpenRouter provider routing preferences. The server applies its zero-data-retention policy last. You cannot turn it off. session_id: type: string maxLength: 256 description: >- Groups related requests in observability tooling. Not sent to the model. trace: type: object description: >- Observability metadata (`trace_id`, `trace_name`, `span_name`, `generation_name`, `parent_span_id`, plus custom keys). user: type: string description: >- A stable id for your end user. The provider uses it for abuse monitoring. example: model: typesafe/jev-1.13 state: ticket: My checkout page shows a blank screen after I click Pay. questions: is_bug: type: noul instructions: Is the customer reporting a software defect? team: type: choice instructions: Which team should own this ticket? criteria: payments: Checkout, billing frontend: Rendering, layout account: Login, permissions urgency: type: score instructions: How urgent is this ticket? criteria: - Can wait for the next release - Should be fixed this week - Blocking revenue right now responses: '200': description: One answer per question, under the question's id. content: application/json: schema: type: object properties: id: type: string example: gen-dec-1790835808-rFEEP39lhplllf50bn9Q model: type: string description: The exact model version that answered. example: typesafe/jev-1.13-20260917 provider: type: string example: TypeSafe answers: type: object description: >- Keyed by your question ids. The fields present depend on the question `type`. additionalProperties: type: object properties: type: type: string enum: - noul - choice - score noul: type: number description: Probability of yes (`noul` questions). choice: type: string description: The selected option id (`choice` questions). score: type: number description: Weighted position on the rubric (`score` questions). probabilities: type: object additionalProperties: type: number legend: type: object description: Level index to rubric text (`score` questions). confidence: type: number description: >- How concentrated the distribution is. It does not say whether the answer is right. usage: type: object properties: input_tokens: type: integer example: 380 output_tokens: type: integer example: 70 cost: type: number description: Provider-reported USD cost. May be 0 or absent. example: id: gen-dec-1790835808-rFEEP39lhplllf50bn9Q model: typesafe/jev-1.13-20260917 provider: TypeSafe answers: is_bug: type: noul noul: 0.94 team: type: choice choice: payments confidence: 0.37 probabilities: payments: 0.58 frontend: 0.42 account: 0 urgency: type: score score: 1.92 confidence: 0.88 legend: '0': Can wait for the next release '1': Should be fixed this week '2': Blocking revenue right now probabilities: '0': 0 '1': 0.07 '2': 0.93 usage: input_tokens: 380 output_tokens: 70 cost: 0 '400': description: >- Invalid request body (`invalid_request`, with `details.validation_errors`). A model that is unregistered, not permitted for your organization, or not a decision model (`model_unavailable`, with `details.reason` one of `not_registered`, `not_permitted`, `not_decision_model`, `proxy_unsupported`). A request the provider rejected (`provider_error`, with the provider's own text in `details.provider_message`). '401': description: Unauthorized. The API key is missing or invalid. '402': description: >- Insufficient credits (`credit_limit_exceeded`), or the provider account is out of credits (`provider_error`). '403': description: Your role does not include agent access (`ACCESS_AGENTS`). '404': description: User profile not found (`user_profile_not_found`). '429': description: >- The provider rate limited the request (`provider_error`). Wait for `Retry-After` seconds before you retry. '502': description: Upstream provider error. security: - bearerAuth: [] x-codeSamples: - lang: bash label: cURL source: | curl -X POST 'https://api.gumloop.com/api/v1/decisions' \ -H 'Authorization: Bearer YOUR_ACCESS_TOKEN' \ -H 'Content-Type: application/json' \ -d '{ "model": "typesafe/jev-1.13", "state": {"ticket": "My checkout page shows a blank screen after I click Pay."}, "questions": { "is_bug": {"type": "noul", "instructions": "Is the customer reporting a software defect?"}, "team": {"type": "choice", "instructions": "Which team should own this ticket?", "criteria": {"payments": "Checkout, billing", "frontend": "Rendering, layout", "account": "Login, permissions"}}, "urgency": {"type": "score", "instructions": "How urgent is this ticket?", "criteria": ["Can wait for the next release", "Should be fixed this week", "Blocking revenue right now"]} } }' - lang: python label: Python source: | import requests response = requests.post( "https://api.gumloop.com/api/v1/decisions", headers={"Authorization": "Bearer YOUR_ACCESS_TOKEN"}, json={ "model": "typesafe/jev-1.13", "state": {"ticket": "My checkout page shows a blank screen after I click Pay."}, "questions": { "is_bug": {"type": "noul", "instructions": "Is the customer reporting a software defect?"}, "team": { "type": "choice", "instructions": "Which team should own this ticket?", "criteria": {"payments": "Checkout, billing", "frontend": "Rendering, layout", "account": "Login, permissions"}, }, "urgency": { "type": "score", "instructions": "How urgent is this ticket?", "criteria": ["Can wait for the next release", "Should be fixed this week", "Blocking revenue right now"], }, }, }, ) answers = response.json()["answers"] if answers["is_bug"]["noul"] > 0.8: route_to(answers["team"]["choice"]) components: securitySchemes: bearerAuth: type: http scheme: bearer description: >- A personal API key or an [OAuth 2.0](/api-reference/oauth) access token. Personal API keys also require the `x-auth-key` header with your user ID. ````