OpenAI made its Decisions API available to all customers on Thursday, Oct. 22, pricing it at $0.10 per million input tokens and $0.01 per million cached input tokens, with no charge for output.
Why it matters: Decision models make quick choices instead of writing full answers. That lets apps and AI agents react in a fraction of a second at a much lower cost.
What it does
The API runs on OpenAI's GPT-6 Luna model. It can classify text and images, evaluate inputs and decide what happens next without generating a full response, OpenAI said in its announcement.
"The big unlock with Decisions API has been the speed, shrinking things from one to three seconds to 100, 120, 180 milliseconds," Nikunj Handa, product lead for OpenAI's APIs, told TechCrunch.
New since the beta
- Native GPT-Live integration so voice apps can react as people speak
- Prompt caching, native JSON inputs and images in question instructions
- Up to 400 questions per request
The competition
The category was created by TypeSafe's Jev model, which made TypeSafe a $7.6 billion company, TechCrunch reported. On public benchmarks, OpenAI's model ranked eighth and fourteenth, behind models from Liquid, Base Lab and TypeSafe.
Early testers included Box, Datadog, Ramp and Speak. Speak's CTO Andrew Hsu said the API cut the time between actions from 1.4 seconds to 0.13 seconds.