Dictionary
Chain-of-Thought Prompting
Instructing a model to think step by step before answering.
1 min readupdated 2026-07-04
/ quick answer
CoT prompting adds 'let's think step by step' or a worked example, boosting accuracy on math, logic, and multi-hop reasoning at the cost of extra tokens. Instructing a model to think step by step before answering.
Instructing a model to think step by step before answering. CoT prompting adds 'let's think step by step' or a worked example, boosting accuracy on math, logic, and multi-hop reasoning at the cost of extra tokens. In practice: 'First list the constraints, then solve.' can lift accuracy by 20+ points on reasoning benchmarks. This dictionary node is part of the Onexial knowledge graph and links to related concepts, workflows and tools below.
Definition
CoT prompting adds 'let's think step by step' or a worked example, boosting accuracy on math, logic, and multi-hop reasoning at the cost of extra tokens.
Example
'First list the constraints, then solve.' can lift accuracy by 20+ points on reasoning benchmarks.
/ frequently asked
What is Chain-of-Thought Prompting?
CoT prompting adds 'let's think step by step' or a worked example, boosting accuracy on math, logic, and multi-hop reasoning at the cost of extra tokens.
What is an example of Chain-of-Thought Prompting?
'First list the constraints, then solve.' can lift accuracy by 20+ points on reasoning benchmarks.
Why does Chain-of-Thought Prompting matter for AI and automation?
Instructing a model to think step by step before answering. It connects to the workflows, prompts and tool stacks linked on this page, so you can move from definition to execution without leaving Onexial.
↳ connected nodes
Dictionary↳ linked
Chain of Thought
Prompting an LLM to reason step-by-step before answering, often improving accuracy on hard tasks.
Dictionary↳ linked
Few-Shot Prompting
Showing the model 2–5 examples of the task inside the prompt so it mirrors the pattern.
Dictionary↳ linked
Zero-Shot Prompting
Asking a model to do a task with no examples in the prompt.
Dictionary↳ linked
Prompt Chaining
Pipelining LLM calls where each step's output feeds the next.
Dictionary↳ linked
System Prompt
A high-priority instruction that sets the model's role, tone and constraints for the whole conversation.
Dictionary↳ linked
Temperature
The randomness knob on an LLM's output distribution.