# OpenAI Decisions API

OpenAI's typed-answer endpoint: probabilities, choices and scores from text or images

**Verdict: conditional.** The Decisions API turns text or images into typed probability, choice or score answers for $0.10 per 1M input tokens with no output charge, and OpenAI says it is about 10x faster than the Responses API. Conditional: it is in public beta, supports only gpt-6-luna, and GA is only expected in the coming weeks. The speed claim is OpenAI's own. It suits teams routing or classifying at volume, not anyone who needs generated text. Last verified 2026-10-29.

The OpenAI Decisions API (POST /v1/decisions) evaluates text, images, JSON or a mix and returns typed answers. OpenAI says it returns typed answers about 10x faster than the Responses API. A request carries a model, an input (a text string, messages with text and images, or structured data passed with an input_json content part) and an array of 1 to 400 questions, each with a type, a name and instructions; images can also sit inside individual question instructions. There are three answer types: predicate (a probability from 0 to 1), choice (the chosen value, a probabilities array and a confidence score) and score (a probability-weighted average with probabilities and confidence). OpenAI lists classification, content routing, request prioritization, damage detection, complaint routing, issue severity assessment and voice-controlled action selection through the Live API as use cases. Only gpt-6-luna is supported. The API entered public beta on October 6, 2026, gained a native integration in GPT-Live on October 15, and became generally available on October 22, 2026; per the changelog, GA also raised the per-request limit from 200 to 400 questions and added prompt-cache discounts for shared input across requests. Uncached input costs $0.10 per 1M tokens, cache hits are discounted at standard gpt-6-luna caching rates (the pricing page lists $0.01 per 1M cached input in short context for gpt-6-luna), and output tokens are not charged. Regional processing premiums and long-context input multipliers apply (the pricing page lists gpt-6-luna input at $0.20 per 1M above 272K input tokens). The API supports Zero Data Retention and HIPAA use for eligible customers, with data residency available in all supported API regions.

## Strengths

- Classifying, routing and prioritizing text, images or JSON where you want a probability, a choice or a score back instead of free text to parse
- High-volume decision calls on cost: $0.10 per 1M uncached input tokens, discounted cache hits and no output-token charge
- Regulated workloads: Zero Data Retention and HIPAA use for eligible customers, with data residency in all supported API regions
- Routing and filtering on thresholds you set: choice and score answers return a probabilities array and a confidence value, and a request can carry up to 400 questions
- Generating or editing text, or running multi-step tool use: it returns typed answers to the questions you ask
- Choosing your model: gpt-6-luna is the only supported model
- Exact cost planning in other regions or on long inputs: regional premiums and long-context multipliers apply, and gpt-6-luna input is $0.20 per 1M above 272K tokens

## Pricing

- Model: paid
- Input tokens: $0.10 / 1M tokens (uncached)

Website: https://developers.openai.com/api/docs/guides/decisions

## Sources

- [OpenAI API changelog, 28 October 2026 entry](https://developers.openai.com/api/docs/changelog)
- [OpenAI Decisions API guide](https://developers.openai.com/api/docs/guides/decisions)
- [OpenAI API pricing, life sciences models](https://developers.openai.com/api/docs/pricing)

Canonical: https://neomanex.com/directory/openai-decisions-api