What makes a good image prompt?

A clear subject with concrete details, one style keyword (photorealistic, oil painting, anime…), one lighting phrase (soft light, golden hour…), no contradictory terms and no vague filler like "beautiful". This rubric checks exactly those things.

Is the score judged by AI?

No. The score is arithmetic: 100 minus fixed deductions from a published 8-check rubric. Every deduction is shown with the rule that triggered it, so you can verify the math yourself.

Why does my prompt lose points for missing style?

Without a style keyword the model picks a look at random — photorealistic, painterly, 3D — and you get inconsistent results. Naming the style is the cheapest way to control output.

Does it work for DALL-E or Flux prompts too?

The vague-word, conflict, length, style and lighting checks apply to any image prompt. The parameter-typo check is Midjourney-specific and simply will not trigger on plain prose.

What is an ai prompt debugger?

An ai prompt debugger is a free online tool that gives you quick, accurate results without spreadsheets or manual math. This version runs entirely in your browser for instant, private results.

How do I use this prompt health debugger & scorer tool?

Enter your details in the fields above and get instant results. Everything runs in your browser — no signup, no waiting, no data uploaded.

Is this prompt health debugger & scorer tool really free?

Yes, completely free with no account required. Use it as many times as you want — there are no usage limits or hidden paywalls.

Prompt Health Debugger & Scorer

What is Ai prompt debugger?

Prompt Health Debugger — Debug your AI image prompt: get a 0–100 health score, find vague words, conflicts, typos and missing style or lighting — with a fix for every issue.

A fixed, fully public 8-check rubric: prompt length (<15 chars −20, >800 −10), vague words from a 17-word list (−5 each, max −15), contradictory term pairs such as dark/bright (−8 each), missing subject detail (−15), missing style keyword from a 20-term list (−10), missing lighting keyword from a 15-term list (−5), known Midjourney parameter typos like --aspec (−5 each) and comma stuffing (>12 commas −8, >20 −12). Score = 100 − deductions, floored at 0; grades A (90+) to F (<40). No AI judgment — pure keyword and string rules, run entirely in your browser.

Why use this ai prompt debugger?

🎯

Built for ai prompt debugger

Debug your AI image prompt: get a 0–100 health score, find vague words, conflicts, typos and missing style or lighting — with a fix for every issue.

📋

Transparent & honest

Keyword lists are fixed English sets — creative or non-English prompts may be mis-scored.

🔒

Private by design

Everything runs in your browser. Your inputs never leave your device.

📱

Works everywhere

Fully responsive — perfect on desktop, tablet, and mobile. No app needed.

🔄

Regularly updated

Regularly maintained and reviewed to ensure accuracy.

💡

Free forever

No accounts, no limits, no watermarks. Use Prompt Health Debugger as much as you need.

How to use Prompt Health Debugger

1Paste your full image prompt, including any parameters like --ar.
2Click Debug to run the 8-check rubric over it.
3Read your score, grade and the plain-language summary.
4Work through each issue — every one includes a concrete fix suggestion.
5Re-paste the improved prompt to watch the score climb.

What can you do with Prompt Health Debugger?

Prompt analyzer — Debug your AI image prompt: get a 0–100 health score, find vague words, conflicts, typos and missing style or lighting — with a fix for every issue.
Fix bad ai prompts — Debug your AI image prompt: get a 0–100 health score, find vague words, conflicts, typos and missing style or lighting — with a fix for every issue.
Prompt quality checker — Debug your AI image prompt: get a 0–100 health score, find vague words, conflicts, typos and missing style or lighting — with a fix for every issue.

Good to know

Keyword lists are fixed English sets — creative or non-English prompts may be mis-scored.
Style/lighting detection is keyword-based; describing style in unusual words will read as "missing".
The score measures prompt hygiene, not artistic quality — a high score does not guarantee a great image.

How it works

A fixed, fully public 8-check rubric: prompt length (<15 chars −20, >800 −10), vague words from a 17-word list (−5 each, max −15), contradictory term pairs such as dark/bright (−8 each), missing subject detail (−15), missing style keyword from a 20-term list (−10), missing lighting keyword from a 15-term list (−5), known Midjourney parameter typos like --aspec (−5 each) and comma stuffing (>12 commas −8, >20 −12). Score = 100 − deductions, floored at 0; grades A (90+) to F (<40). No AI judgment — pure keyword and string rules, run entirely in your browser.

Examples

Thin vague prompt

Scores around 60 (C): flags "beautiful" as vague, plus missing style and lighting, each with a fix.

Typo + conflict

Flags the dark/bright conflict and corrects --aspec to --ar.

People also search for these — this tool handles them all:

ai prompt debuggerprompt analyzerfix bad ai promptsprompt quality checker

Frequently asked questions

Explore more free tools in AI Tools: