Score and fix any prompt, free
Paste a draft below. Get a 0-100 score, the weak spots flagged, and a cleaner rewrite. No signup. 3 free runs per day.
Want to build a prompt from scratch? Open the Guided Builder
How it works
Paste your draft
Drop in the prompt you are about to send to ChatGPT, Claude, Gemini, or any model. A rough first try is exactly what this is for.
Read the score and gaps
Get a 0-100 rating in seconds plus the specific reasons it underperforms: missing role, undefined output format, buried instruction, no success criteria.
Copy the rewrite
Take the optimized version and get a sharper answer the first time. Each fix is explained, so your own prompts get stronger over time.
What you get
A real score
Every prompt gets a 0-100 grade against a consistent rubric, so you know whether it is ready or still leaking quality.
The weak spots, named
It points at the exact lines costing you: the vague ask, the missing format, the instruction the model will skip.
A rewrite you can ship
You get an optimized version restructured around what the model rewards, not a list of tips you have to apply yourself.
No signup to start
Paste a prompt and get your score and rewrite with no account and no credit card. 3 free runs a day.
Common questions
- Is it actually free?
- Yes. You get 3 free QA reports a day with no account and no credit card. Paste a prompt and you get the score and the rewrite.
- Which models does it work with?
- The guidance targets the models most people use, including ChatGPT, Claude, and Gemini. The rewrite focuses on prompt structure, which carries across models.
- What does the score measure?
- It grades your prompt against a rubric covering role, task clarity, output format, context, and success criteria, the elements that decide whether a model answers well in one pass.
- What do I get if I sign up?
- A free account adds saved history, more daily reports, deeper Sonnet and Opus analysis, image-prompt QA, and variant rewrites.