Socratic prompting
Reasoning
Level 2 · Advanced reasoningThe model asks clarifying questions before attempting to solve the problem, instead of answering immediately.
Socratic prompting asks the model to identify what's missing or ambiguous about the request and ask a small number of clarifying questions, instead of answering immediately with whatever assumptions it would otherwise make silently. It trades one extra round-trip for a much better-targeted final answer, especially useful when the task is underspecified.
When to use it
Use it for genuinely underspecified tasks, where guessing the missing details would likely produce an answer that misses what you actually needed — early-stage planning, ambiguous requests, requirements gathering.
When to avoid it
Avoid it for well-specified tasks — asking clarifying questions when nothing is actually missing just adds an unnecessary round-trip. Also avoid it in non-interactive contexts where there's no way to answer the follow-up questions.
Advantages
- Surfaces missing information before an answer is built on a wrong assumption
- Particularly useful for interactive use, where a follow-up round-trip is cheap
Disadvantages
- Doesn't produce a final answer in a single turn — requires the user to respond to the questions
- Not useful in a single-shot / non-interactive context
Examples
Help me plan a marketing campaign for a new product launch. Before proposing anything, identify the most important missing information and ask up to 5 clarifying questions.
Pairs well with
Try it
Fill in the placeholders below and see the compiled prompt update live.
What you want to achieve.
Maximum number of clarifying questions to ask (default 5).
- 17
- Words
- 154
- Characters
- 0/2
- Placeholders filled
- 1 min
- Reading time
- Completeness0
- Clarity0
- Context0
- Specificity0
- Constraints0
- Format0
- Output0
- Constraints note0
- Success criteria0
- Scope focus100
- Verification step0
- Negative example0
- Confidence0
- Placeholder syntax100
- Markdown structure0
- Fill in the remaining placeholders.
- Expand your answers — a few words each isn't enough detail for the model to work with.
- Add background context so the model understands the situation.