# What is Reasoning Model? Definition & Technical Mechanics

> **Source:** [SPIDITS AI AI Glossary](https://spidits.com/ai-glossary)  
> **Category:** Foundational AI  
> **Canonical Citation:** [https://spidits.com/ai-glossary/reasoning-model](https://spidits.com/ai-glossary/reasoning-model)

## Definition
A Reasoning Model (or o1-style model) is an artificial intelligence model trained to perform reinforcement learning and execute chain-of-thought steps internally before returning an answer. This allows the model to deliberate, correct mistakes, and evaluate strategies.

## Architectural Mechanics & Deep Dive
A reasoning model is a class of Large Language Models specifically trained to perform explicit, step-by-step thinking (system 2 thinking) before generating outputs. By incorporating a hidden thinking phase where the model self-corrects, plans, and decomposes problems, reasoning models achieve high accuracy in mathematics, coding, and logic.

## Industry Applications & Real-World Use Cases
Complex code generation, advanced mathematics solving, and scientific research analysis.

## Frequently Asked Questions

### How do reasoning models differ from standard LLMs?
Standard LLMs output text instantly on a token-by-token basis. Reasoning models generate internal 'thinking tokens' to plan and check their work before writing the final output.

### Do thinking tokens count toward API costs?
Yes, thinking tokens require computation during inference and are typically charged as input/output tokens.

## Related Terminology
- [llm](https://spidits.com/ai-glossary/llm)
- [chain-of-thought](https://spidits.com/ai-glossary/chain-of-thought)
- [reinforcement-learning](https://spidits.com/ai-glossary/reinforcement-learning)

---
*From the SPIDITS AI Knowledge Base (336+ terms indexed). Explore full technical definitions: [https://spidits.com/ai-glossary](https://spidits.com/ai-glossary)*