What is chain-of-thought prompting?
Also called: chain of thought, step-by-step reasoning
Chain-of-thought prompting asks an AI model to work through a problem step by step before giving the final answer. Writing out the intermediate steps often makes answers to maths, logic and multi-step questions more accurate, and makes mistakes easier to spot.
How does chain-of-thought prompting work?
You add an instruction like 'Think it through step by step' or show an example with worked steps. Many newer models also reason internally before they answer, which is a built-in form of the same idea.
Why does chain-of-thought prompting matter?
It improves accuracy on tricky problems and lets you check the reasoning instead of trusting a bare answer.
An example
Example: Asked to split a ₹2,340 restaurant bill with a 10% service charge among four friends, the model shows each step, so you can check the maths.
Related terms
Learn it properly
Chain-of-thought prompting is covered in Use AI Well, with short lessons, practice questions and instant feedback.