What it does

The Decisions API is a dedicated endpoint, POST /v1/decisions, that evaluates text, images or both and returns typed answers instead of free text. You send shared input plus one or more named questions and get back an answer for each. There are three question types. A predicate gives the probability that a condition is true, a choice selects one of your supplied options and reports a probability for each, and a score rates the input against ordered levels as a probability-weighted average. The page says these answers arrive about ten times faster than the same task through the Responses API, and suggests using them to classify content, route requests and prioritise work. For generating objects that follow your own schema, it points to Structured Outputs instead. The API is in public beta, with general availability expected in the coming weeks, and gpt-6-luna is the only model available. Images must be sent as inline base64 data URLs; hosted image URLs and file IDs are not supported. The page gives examples for the Python, JavaScript, Go, Ruby and Java SDKs.

Capabilities

Predicate questions return a 0 to 1 probability that a condition is true
Choice questions pick from options you define and return probabilities
Score questions rate input against ordered rubric levels
Accepts text, images or both in one request
SDK examples for Python, JavaScript, Go, Ruby and Java

When to use it

You need to classify content or route support requests in code
You want a probability to set a threshold on, such as flagging product photos with visible damage
You want to rank items against a rubric, such as issue severity