Score any AI prompt in seconds
Already have a prompt? Paste it below to rate it across seven research-backed dimensions, and get focused tips on how to make it stronger. Free, instant, no signup.
What a high score looks like
Here is a real prompt run through the same scorer. Notice how every dimension is covered, not just one or two.
You are a senior content strategist with 10 years of experience writing for speciality coffee brands. Write a 1,200-word blog post about pour-over brewing for home baristas in the UK who already own a gooseneck kettle. The target audience is curious hobbyists, and the goal is to convert readers into subscribers to our monthly bean club. Structure the post with one H1, five H2 sections, and a short FAQ. You must keep each paragraph under 80 words. Use a warm, plain-spoken tone and British English. Only use jargon if you explain it: for example, describe the bloom as 'the first pour that lets trapped gas escape'. Do not include affiliate links.
Want to see a weak prompt turned into this one, step by step? Read what makes a high-scoring prompt.
What the scorer checks
Seven dimensions, weighted by how much each one affects output quality. For each, here is the difference between a low score and a high one.
Role definition
12%Low: Help me with marketing.
High: You are a senior B2B performance marketer.
Task clarity
18%Low: Look at my landing page.
High: Rewrite my landing page headline for clarity, then list three alternatives.
Specificity
20%Low: Write a few social posts.
High: Write 5 LinkedIn posts of 80 to 120 words for UK SaaS founders.
Context
15%Low: Write a product description.
High: Write a product description for a 35 GBP candle sold by a small UK brand.
Output format
15%Low: Compare these three tools.
High: Compare these three tools as a markdown table with Price and Best for columns.
Constraints
10%Low: Write a CEO update.
High: Write a CEO update under 250 words. No buzzwords. British English.
Examples
10%Low: Write taglines in our voice.
High: Write taglines in our voice. For example: 'Boring is the new bold.'
For the full walk-through of each dimension, read the prompt engineering guide.
Frequently asked questions
How do I score a prompt?
Paste your prompt into the box above and the scorer rates it instantly across seven dimensions: role, task clarity, specificity, context, output format, constraints, and examples. You get an overall score out of 100 plus a focused tip for each dimension that is pulling the total down. Scoring runs entirely in your browser, so your prompt never leaves your device.
Is this prompt quality checker free?
Yes, completely. There is no signup, no usage limit, and no paywall. The scorer is free to use as often as you like, and the optional AI rewrite is free too. The site is funded by display ads on the surrounding pages.
What counts as a good prompt score?
Anything 85 or above is excellent and should produce strong, focused output. 70 to 84 is good with room to sharpen. Below 50 means the model is having to guess at most of what you want. Treat the score as a directional signal of what is missing rather than an absolute verdict; the real test is whether the prompt gets you the output you need.
Does the scorer work for ChatGPT, Claude, and Gemini?
Yes. The seven dimensions are model-agnostic, so a prompt that scores well here will perform better on ChatGPT, Claude, Gemini, Llama, Mistral, and any other capable large language model. The scoring is about how clearly you have briefed the model, not about model-specific syntax.
Can I rate a prompt without writing one first?
If you do not have a prompt yet, use the guided builder on the homepage. Answer a few questions and it assembles a high-scoring prompt for you, which you can then paste back here to see the breakdown.
Don't have a prompt yet?
Try the guided prompt builder on the homepage: answer a few questions and we'll write a high-scoring prompt for you.
Open the prompt builder →