OpenAI answers TypeSafe’s Jev with a Decision API built on Luna

With the sudden rise of TypeSafe’s Jev, decision models have become incredibly popular, so it’s not surprising that OpenAI is also announcing its take on this model type at its annual DevDay conference on Tuesday.

The company’s new Decisions API is based on the company’s Luna model — the smallest and most affordable model in its current lineup.

Limited preview

The Decisions API is likely a reaction to TypeSafe and Jev, and OpenAI probably rushed the announcement ahead of its DevDay, so for now, this is all OpenAI is sharing about the Decisions API.

An OpenAI spokesperson tells The New Stack that the company plans to share more “at broad rollout.”

For now, the new API is available in limited preview, and the broad release is planned for the coming days.

Predefined answers, real confidence scores

The core idea behind these decision models is that they are explicitly not chat models; instead, they return a set of predefined answers with confidence scores. Regular LLMs are not always very good at this. Their confidence scores, after all, are often a rough guess — and they burn quite a few tokens to get there.

This makes this kind of model ideal for classifying content, routing requests, or choosing an agent’s next action from a limited set of choices.