Reasoning Model
A model optimized for complex problem solving by generating a longer internal reasoning process before answering.
In plain language
Reasoning models are trained with reinforcement learning or large reasoning datasets and perform better on math, code, planning and multi-step judgment. They spend more tokens thinking, so latency and cost are higher. They fit tasks with verifiable answers and multiple steps, not simple classification, rewriting or low-latency chat. Their stated reasoning is not necessarily the actual decision path and should not be treated as proof of interpretability.
Example
A reasoning model is usually more reliable for multi-step math or reviewing concurrent code, but using one to tweak a sentence wastes cost.
Recommended tools
Related terms
Chain of Thought
Prompting a model to write intermediate reasoning steps before concluding.
Inference
The computation a trained model performs to produce output from an input.
Test-Time Compute
Spending more computation at inference time—longer thinking, multiple samples or search—to improve answer quality.
Agent Memory
A mechanism for an agent to store and retrieve user preferences, past tasks and key facts across sessions.