---
title: "What is chain-of-thought prompting?"  
description: "What is chain-of-thought prompting?"  
author: "Alice Bertolini"  
published: 2026-11-15  
updated: 2026-11-15  
canonical: https://answers.mindstick.com/qa/120087/what-is-chain-of-thought-prompting  
category: "Prompt Engineering"  
tags: ["prompt engineering", "Reasoning", "techniques"]  
reading_time: 3 minutes  

---

# What is chain-of-thought prompting?



## Answers

### Answer by Grace Kelly

Chain-of-thought prompting is a technique where you ask the AI to work through a problem step by step before giving its final answer, rather than jumping straight to a conclusion. Researchers found in 2022 that this simple change significantly improved AI accuracy on maths, logic and multi-step reasoning tasks.

The classic version is simply adding a phrase like "Let's think step by step" or "Show your reasoning before giving the final answer."

Why does it work? Language models generate text one piece at a time. When an AI writes out intermediate steps, each step becomes part of the context for the next one, much like a student writing rough work on paper. This helps the model break a complex problem into manageable parts and reduces careless errors.

Here is an example.

Without chain of thought: "A shop sells pens at 12 rupees each. If you buy 15 pens, you get a 10 percent discount on the total. How much do you pay?" The AI might give a quick answer and occasionally get it wrong.

With chain of thought: "Solve this step by step: first calculate the total before discount, then the discount amount, then the final price." The AI writes 15 × 12 = 180, 10 percent of 180 = 18, 180 − 18 = 162 rupees. The steps make the answer more reliable and easy to check.

When is chain of thought useful?

1. Maths and calculations.\
2. Logic puzzles and reasoning questions.\
3. Complex decisions with several factors.\
4. Analysing documents to reach a conclusion.\
5. Debugging code.\
6. Any task where you want to see how the AI reached its answer.

How have reasoning models changed this? Modern reasoning or "thinking" models, available in ChatGPT, Claude, Gemini, Grok and others, are trained to reason internally before answering. With these models, you usually do not need to say "think step by step," and overly detailed instructions on how to think can even get in the way. A clear description of the problem is enough.

Chain-of-thought prompting is still useful with standard models, and asking for visible steps remains valuable when you want to check the reasoning, teach someone the method or audit the answer.

One caution: showing reasoning steps does not guarantee correctness. A wrong step can lead to a confident wrong answer. Check the steps, especially for important calculations or decisions.


---

Original Source: https://answers.mindstick.com/qa/120087/what-is-chain-of-thought-prompting

Copyright © MindStick Software Pvt. Ltd. This Markdown version is provided for developers, AI systems, and offline reading.