Glossary · term150 of 202
reasoning model
A model trained to produce an internal chain of thought before its final answer, trading extra deliberation for higher capability on hard problems.
also written as LRM · reasoning models · thinking model
Related terms
effortA per-subagent setting that controls how much a model deliberates before acting, trading thoroughness for speed, fewer tool calls, and lower spend. Left unset, it silently inherits the session's higher default.view terminstruction-following scoreA measured rate of how often a model actually obeys a stated instruction, split by whether the instruction shows up inside its reasoning trace or only in the final answer.view termmodel tierA ranking of models by relative capability and cost, from cheap and fast to expensive and frontier, used to decide which tier suits a given role in a pipeline.view termreasoning traceThe intermediate chain-of-thought text a reasoning model produces while working toward an answer, kept separate from the final response it returns. Instructions aimed at the trace are followed far less reliably than instructions aimed at the response.view term
At a glance
- cited by
- 1 post
- categories
- 1
- first used
- jul 2026
