# Reasoning model · Glossary
A language model trained to work through intermediate steps before producing an answer, typically spending more computation at inference time on harder problems.
[Glossary](/glossary) · Agentic systems and generative AI

# Reasoning model

     A language model trained to work through intermediate steps before producing an answer, typically spending more computation at inference time on harder problems. Reasoning models tend to perform better on multi-step analysis, planning and coding tasks.

## Why it matters

     Reasoning models can make agents and analytical assistants more capable, but they are usually slower and more expensive per request, and their extra steps do not guarantee a correct or well-grounded answer.

     Choosing a model should follow evaluation on the organisation’s own tasks. Many production workflows combine a reasoning model for difficult steps with faster, cheaper models for routine classification or extraction.

## In practice

     For example, an energy trading analytics team might use a reasoning model to draft a structured explanation of an unusual price movement from several data sources, while a smaller model handles routine tagging of market news.

## Where Rodan fits

     Rodan selects and combines models based on measured task performance, cost and latency in [Applied AI Engineering](https://rodan.io/applied-ai-engineering) work. See also [how to measure AI performance beyond accuracy metrics](https://rodan.io/insights/how-to-measure-ai-performance-beyond-accuracy-metrics).

## Related terms

- [Large language model](/glossary#large-language-model)

- [Inference](/glossary#inference)

- [AI evaluation (evals)](/glossary#ai-evaluation)

- [Small language model (SLM)](/glossary#small-language-model)

- [Token](/glossary/token)
HTML: https://rodan.io/glossary/reasoning-model