Decision models are suddenly everywhere. OpenAI’s is now public.
This week, OpenAI opened its Decisions API to public beta and added image inputs to make fast decisions from visual The post Decision models are suddenly everywhere. OpenAI’s is now public. appeared first on The New Stack .
OpenAI has recently opened its Decisions API to a public beta, allowing developers to turn text or images into structured outputs such as predicates, choices, and scores. This move comes in response to TypeSafe's Jev, and the API is now being billed at $0.10 per 1M input tokens, with no charges for output tokens, cache reads, or writes.
The API, running on GPT-6 Luna, can handle three types of outputs - predicates (likelihood of something being true), choices (best fit among predefined options), and scores (numeric range). OpenAI has demonstrated the API's capabilities with various examples, including a video game scenario where it can decide if a car should switch lanes or keep driving.
The company also hinted at potential applications, like taking screenshots and quickly choosing the next action. Perplexity, Cloudflare, and Amazon have also released their own decision models this month, with open-weight models that developers can download and run on their own infrastructure. These models include Perplexity's pplx-decider-v1-27b, Cloudflare's Clef and Clef-flash, and Amazon's Strands Decider 2B, all of which can accept both text and image inputs.
OpenAI, however, has opted to keep its model behind a hosted API, which may become a crucial factor in developers' decision-making process as decision models become more prevalent.
Written by urgent.news from The New Stack's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
