mediumAI Engineering

What is chain-of-thought prompting and when does it actually help?

Q: What is chain-of-thought prompting and when does it actually help?

**TL;DR.** Chain-of-thought (CoT) asks the model to **reason step by step before answering**. It lifts accuracy on math, logic and multi-step tasks because each reasoning token conditions the next; it wastes tokens on simple lookups. **Reasoning models** now internalize it with dedicated thinking tokens. direct question answer (guess) CoT <rect class='d-box' x=

481 views

Understand the problem

Making the model 'think out loud' — why it works, when it is wasted tokens, and reasoning models.

chain-of-thoughtreasoningprompting

Attempt it yourself

Sketch your approach before reading the solution — that's what interviews test.

Stuck? AI Nudge Available

Get a conceptual hint to guide your logic without spoiling the final implementation.

Study the solution

The solution is waiting

Give it an honest attempt first — then compare your thinking with the full walkthrough.

Read the code

Structured reason-then-answer output

const prompt = [
  "Solve the scheduling question below.",
  "First write your reasoning inside <thinking> tags.",
  "Then output ONLY the final schedule as JSON inside <answer> tags.",
  "",
  question,
].join("\n");

const reply = await complete(prompt);
const answer = between(reply, "<answer>", "</answer>"); // parse the fenced part only

Join the discussion

Discussion (0)

No responses yet. Be the first to share what you think.

What is an AI agent and how does the agent loop work?

How do you choose between open-weight models and proprietary API models?