Provide the context
Send the incoming request, a generated draft, or the records needed to answer your decision questions.
state = ticket.textDecisionsApi · Decision API OpenAI
Add an independent decision layer to applications that use OpenAI. Test how to classify requests, score urgency, and route the next step with structured answers your backend can use.
Decision workflow example
No credits usedCustomer message
“I was charged twice this morning. Please refund the duplicate payment today.”
One message → three decision signals
Choice
Billing
Select the billing workflow
Yes / no
96%
Probability of an urgent request
Score
2.8 / 3
Urgency on a 0–3 scale
Static illustration of this platform's output format. These values are not a live response, an accuracy claim, or an OpenAI API response.
01 / Workbench
Try OpenAI Decisions API with GPT-6 Luna, or choose another decision model. Edit the input, define typed questions, and inspect the results before connecting the workflow to your backend.
Live requests require sign-in and use your available account credits.
02 / Connect the workflow
Your backend can call a decision model before or after an OpenAI request. Supply the relevant context, inspect the result, and apply your own rules before taking action.
Send the incoming request, a generated draft, or the records needed to answer your decision questions.
state = ticket.textSpecify the routing labels, scoring levels, or yes-and-no criteria that your application understands.
questions = { intent, urgent }Read each typed answer alongside its probability signals and request details before using it in your workflow.
answer = result.answers.routeUse the result to select a workflow, continue a model call, or request human review. Your application controls execution.
if result.urgent: escalate()03 / Practical applications
Start with one repeated judgment in your product. Test representative and ambiguous inputs before connecting the result to an automated action.
Assign billing, technical, account, or other labels before choosing a support workflow or preparing context for a generated reply.
Decide whether a request needs a record lookup, an OpenAI model call, another tool, or human review using options you define.
Evaluate a draft against explicit relevance or completeness criteria, then use the score to send selected responses for further review.
04 / Work with typed outputs
Express classification, scoring, and focused checks as bounded questions. The returned signal gives your application a defined input for its next step.
Choose from labels you provide, such as billing, support, or review. Map each answer to an existing application workflow.
Define ordered levels for urgency, relevance, or completeness. Evaluate the same criteria across examples from your OpenAI application.
Ask a focused question and use its returned probability with a threshold you have evaluated on your own cases.
Integrate from your backend
The example below calls this platform's /v1/systemone endpoint with Jev, using a platform API key. Keep it as a separate server-side request in your workflow; this is not an OpenAI endpoint or an OpenAI SDK example.
Read the platform API docsThis platform's decision API
Server-side JavaScript · Jev request example
const response = await fetch(
"https://decisionapi.net/v1/systemone",
{
method: "POST",
headers: {
"Authorization": `Bearer ${DECISIONS_API_KEY}`,
"Content-Type": "application/json"
},
body: JSON.stringify({
model: "typesafe/jev-1.14",
state: ticket,
questions: {
route: {
type: "choice",
instructions: "Choose the support queue for this ticket.",
criteria
}
}
})
}
);05 / Understand the service
Create an account, check your available testing credits, and try real classification and scoring cases before choosing a credit pack.
Use the same credit balance across the playground, API, and evaluation workflows. Credits from purchased packs do not expire.
Review request history and credit activity, and use request identifiers to connect API calls with their usage records.
Measure credit usage with representative requests, then choose a one-time pack. Credits pay for GPT-6 Luna, Jev, and other calls made through this platform, separately from direct OpenAI billing.
One-time plans do not auto-renew or require a long-term subscription; the $10, $100, and $1,000 tiers include credits that remain available after purchase.
Model access and billing
OpenAI’s Decisions API is in public beta. This platform provides model access through its unified API; check the current directory and your account access for available models and question types. Credit packs fund this service and do not add credit to an OpenAI account.
Starter
Validate one real workflow, from the playground to your first API call
Pro
Connect classification, routing, and safety decisions to a production product
Enterprise
For multiple workspaces, team collaboration, and custom production integrations, with 10% extra credits
Compare before you ship
Use the same human-labeled cases to compare mistakes, human-review rates, response times, and credits used. Include clear, ambiguous, and unusual inputs, then choose the model that fits your workflow.
Jev
Typed choices, scores, and yes–no decisions
Laya
English and multilingual decision workflows
Kev, Solar, and Span
Additional models for comparison and evaluation
OpenAI Decisions API
GPT-6 Luna for structured decision workflows
Recent blog posts
Product notes, technical details, and practical ways to use Jev.

AI APIs
Compare Decisions API vs Jev: platform and model differences, typed outputs, API integration, cost evaluation, and a practical checklist for choosing your stack.
Oct 3, 2026 · 11 min read

AI Models
Understand the OpenAI Decisions API Luna model, preview status, bounded decisions, routing contracts, evaluation, and a practical path to production.
Oct 3, 2026 · 12 min read

Developer Guide
Build a decision API OpenAI workflow with typed outputs, a REST example, evaluation metrics, and production controls. Understand the different API contracts.
Oct 3, 2026 · 11 min read

AI APIs
A practical Decisions API documentation guide with the REST request format, Choice, Score, and Noul questions, a support-routing example, and production safety patterns.
Oct 1, 2026 · 11 min read
Understand how an independent decision layer can fit into your OpenAI application workflow.
No. This is an independently operated platform for decision models. This page describes how your application can use its decision endpoint within a workflow that also uses OpenAI; it does not claim an official OpenAI service or affiliation.
The plans and credits on this page apply to this platform's services. They do not include or replace billing for requests you send separately to OpenAI. Check each service's usage and pricing for your workflow.
The example uses typesafe/jev-1.14 through this platform's /v1/systemone endpoint. Use the playground and documentation to inspect the decision models and question types available for your task.
Send the information needed for the judgment: an incoming message, relevant records, evaluation criteria, or a draft from your application. Keep the question focused and include the context required to interpret it.
No. Probability is a model signal, not a guarantee. Evaluate it on representative cases, define thresholds for your workflow, and keep a review path for uncertain or consequential decisions.
Create a platform API key, send state and questions from your backend, and inspect the structured response. Your application can use the result to route a subsequent OpenAI request or review an earlier output. Follow this platform's API documentation for the decision call.