Free tool. No signup. Runs in your browser.
AI Prompt Optimizer
Paste a prompt, get a score out of 100 and a rewritten version that ChatGPT, Claude and Gemini answer better.
Paste exactly what you would type into ChatGPT, Claude or Gemini. Nothing leaves your browser.
Scoring is rule-based and runs on this page. It checks the nine things that most often decide whether a model answers well on the first try; it does not send your text to any AI service.
Your score and rewrite appear here
Nine checks: task, context, audience, format, role, constraints, example, definition of done, specific wording.
How to use the Prompt Optimizer
- 1
Paste your prompt
Use the exact text you were about to send. Longer is fine; the checks care about what is present, not how it is phrased. - 2
Read the checklist
Each of the nine checks shows what the analyzer found and one concrete fix. Missing items cost the most, partial ones cost half. - 3
Copy the rewrite
The improved prompt keeps your sentences and adds bracketed slots. Fill them, delete what does not apply, and send.
Next step
Learn to write prompts that work the first time
Score a prompt above and this block names the checks the course would fix. Coursiv's ChatGPT and Claude courses teach the method behind all nine.
A short quiz first. No payment details.
Why most prompts score under 50
Most prompts are a task with nothing around it: "write a post about our new feature". The model has to guess who the post is for, how long it should be, what tone the company uses and what would make the result usable. It guesses the average answer, and the average answer is what people call generic AI writing.
The fix is not a longer prompt, it is a more complete one. Adding an audience, a format with a length, one constraint and one line about what a good result must do typically moves a prompt from the 30s to the 80s on this scale, and that difference shows up as fewer follow-up rounds.
The nine checks, and how much each is worth
Clear task (20 points), context (15), output format and length (15), audience (10), constraints and tone (10), role (8), example (8), definition of done (8), and specific wording (6). The weights reflect how often a missing item causes a wasted first answer: a missing task or format nearly always does, a missing role usually does not.
Role and example are scored as partial rather than missing when absent, because many everyday prompts do not need them. The rewrite still leaves a slot for each, so you can decide.
Prompt optimizer, enhancer, improver or builder?
The names overlap. A prompt enhancer or improver rewrites your text; a prompt builder assembles one from fields; a prompt optimizer scores it. This tool does all three in one pass: the score tells you how far you are from a first-try answer, the checklist tells you why, and the rewrite is a builder pre-filled with your own words.
More free AI tools
Token Counter
Count the tokens in any text for GPT, Claude and Gemini models and see what a request, or a month of requests, costs.
Open tool →Hook Generator
Turn one topic into 20 opening lines built on proven hook patterns for LinkedIn, X, Instagram, YouTube and email.
Open tool →Formula Generator
Describe what you need in plain words, fill in your ranges, and copy a working formula for Excel or Google Sheets.
Open tool →