Glossary
Reasoning model
A model trained to work through a problem in intermediate steps before giving a final answer.
A reasoning model is a language model trained to spend extra computation working through a problem, often by producing a hidden or visible chain of intermediate steps, before it answers. This tends to help on math, coding, and multi-step planning.
The trade-off is cost and speed: more "thinking" means more tokens and more inference computing per answer. Many products let users or developers choose how much reasoning to allow.
The word "reasoning" describes a technique and a product category. It does not settle debates about whether the model reasons the way people do, and the steps a model displays may not faithfully reflect how it reached its answer.