Techniques6 min read
Chain-of-Thought Prompting Explained: Does It Still Work?
Chain-of-thought prompting asks a model to reason step by step before answering. See when it helps, when reasoning models make it redundant, and how to use it.
Reasoning
Chain-of-thought and reasoning prompts, explained
Reasoning prompts ask the model to work through a problem step by step before giving its final answer, an approach known as chain-of-thought prompting. It helps on multi-step math, logic and planning, and matters less for simple lookups. Modern reasoning models already think internally, so the skill today is deciding when to ask for visible reasoning and how to keep it separate from the final answer.
Coming soon onApp StoreComing soon onGoogle Play
Structured outputIntermediate
Your app parses the model’s JSON, but sometimes the reply starts with “Sure! Here is the JSON:”. What is the most robust fix?
AAdd “please” to the prompt
BUse structured-output mode with a schema, then validate
CLower the temperature to 0
DStrip the first line of every reply
Sample question · Reasoning
You need a model’s reasoning for debugging but only the final number in your app. What prompt structure fits best?
- AAsk for the answer only
- BAsk for reasoning inside <thinking> tags, then the answer inside <answer> tags
- CAsk for the answer first, then the reasoning
- DAsk it to be confident
Show the answer and why
B. Ask for reasoning inside <thinking> tags, then the answer inside <answer> tags Reasoning written before the answer can improve it, and tags let your code extract only the final answer while you keep the reasoning for logs.
What the questions test
- When step-by-step reasoning improves results
- Reasoning before the answer, not after
- Breaking a big task into a chain of prompts
- Asking a model to check its own work
What you’ll learn
- What chain-of-thought is and where it helps
- How to separate reasoning from the final output
- How to split complex tasks into steps
- How to use self-checks without wasting tokens
How to practice it in TokIQ
- 1. Open Reasoning. Pick it from the topic list on the Quiz tab. It is a Premium topic; the free plan includes Fundamentals, Role & context and Examples.
- 2. Answer one question at a time. You see right away whether you were correct.
- 3. Tap Why? Read the short explanation and answer three follow-up questions on the same idea.
- 4. Watch your accuracy. The Progress tab shows how Reasoning compares with your other topics.
Reasoning: common questions
What is chain-of-thought prompting?
Chain-of-thought prompting asks the model to write out intermediate reasoning steps before its final answer. It was popularized by Wei et al. (2022) and helps most on multi-step problems.
Should I still write “think step by step”?
With reasoning models it is often unnecessary. With other models it can still help on multi-step tasks, especially if you ask for the reasoning before the answer and in a separate section.
Further reading
Learning6 min read
How to Learn Prompt Engineering: A Realistic Step-by-Step Path
A realistic path for learning prompt engineering: what to learn first, which free resources are worth it, how to practice, and how to know you're improving.
Basics6 min read
What Is Prompt Engineering? A Practical Guide With Examples
Prompt engineering is the skill of writing inputs that get reliable results from AI models. Learn the parts of a good prompt, with before-and-after examples.
More topics
Practice reasoning a few questions a day.
Short quizzes on real prompting decisions, with an explanation for every answer. Free to start on iPhone and Android.
Coming soon onApp StoreComing soon onGoogle Play