Decision Model
A model built to answer fixed questions with a choice and a confidence score, instead of writing text.
Definition
A decision model is a kind of AI model built to decide rather than write. Instead of generating an answer word by word like a chatbot, you hand it some content plus a set of questions with fixed possible answers — is this spam, which category does it belong to, rate this one to five — and it returns the choice along with a confidence number for each option. Because it never writes prose, there is nothing to parse and no way for it to wander off-format, and it answers in milliseconds rather than seconds. The approach suits high-volume, repetitive judgments like moderation, routing, and filtering, where a full language model is slower and costlier than the job needs. Research papers and model registries also call these System One models, after the psychology idea of fast intuitive judgment. The category took shape in late 2026 around models such as Jev, Clef, and Kev, and has since grown to include entries from a range of companies, built either from scratch or as a decision layer bolted onto an existing open model.