Prompt score
Waiting for analysis…
Paste a prompt, pick the model you'll run it on, and get it linted & rewritten against that vendor's own guide — Anthropic, OpenAI, Google, Perplexity. Free, open source, 100% in your browser.
Waiting for analysis…
Rules-based rewrite: your prompt in the vendor's preferred shape, with [TODO] markers where the guides say something's missing. Connect your own AI below for a fully written rewrite.
Use any model you have a key for — the big four, or any OpenAI-compatible endpoint (OpenRouter, Groq, Mistral, DeepSeek, xAI, local Ollama…). Calls go straight from your browser to the provider; this site has no server. Some providers may block browser calls (CORS).
Your prompt is checked against rules distilled from each vendor's docs. Every finding links to its source.
Free, deterministic rewrite into the target's preferred shape — XML tags, context-first, or search-tight.
Optionally hand it to a real model with your own key for a fully AI-written rewrite. No server involved.