# Overthinking

> Overthinking is the tendency of reasoning models to spend far more reasoning tokens than a problem requires, for example writing long chains of thought for trivial arithmetic, which adds cost and latency with little or no accuracy gain.

- Identifier: PTL-0170
- Category: Failure Modes & Evaluation
- Canonical URL: https://protologue.com/t/overthinking/
- Also known as: overthinking problem, excessive reasoning
- Introduced: 2024

## Description

Analysis of o1-style models found that later reasoning rounds on simple problems rarely contribute new solutions. Remedies include adjustable reasoning effort or budgets, routing easy queries to non-reasoning models, and training that rewards efficient reasoning.

## Related terms

- [Reasoning Model](https://protologue.com/t/reasoning-model/)
- [Extended Thinking](https://protologue.com/t/extended-thinking/)
- [Budget Forcing](https://protologue.com/t/budget-forcing/)
- [Chain of Draft](https://protologue.com/t/chain-of-draft/)
- [Test-Time Compute Scaling](https://protologue.com/t/test-time-compute-scaling/)

## Sources

- Chen et al. (2024). Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs. https://arxiv.org/abs/2412.21187

## Cite this entry

Protologue. (2026). Overthinking. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.4.0, PTL-0170). https://protologue.com/t/overthinking/

License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)