About this project
What it is, why it exists, and how it works.
Why this exists
Search is shifting. More people get their answers directly from ChatGPT, Perplexity, or Gemini — without clicking a link. Traditional SEO tools tell you how you rank in Google. They don't tell you whether an AI will cite you.
This scanner fills that gap. It checks the signals that determine AI visibility: structured data, bot access policy, content clarity, entity consistency, and more. It also lets you fire real prompts at live AI search engines to measure your actual citation rate.
The tool is open to anyone. Basic scans are free. AI visibility tests cost credits because they make real API calls.
Who built it
Built to explore what AI-age website hygiene actually looks like in practice — and to help website owners, marketers, and developers understand their real AI visibility.
info@llmzap.comHow it works
The scanner fetches the target URL plus a set of side
requests: robots.txt, llms.txt, sitemap.xml,
/.well-known/tdmrep.json, and /pricing.md.
It parses JSON-LD blocks, detects the page type, and takes a headless-browser
screenshot to evaluate render fidelity.
12 check functions run against the fetched data — covering SEO, search presence, AI bot policy, structured data, content quality, E-E-A-T, llms.txt, rights & licensing, entity understanding, hallucination resistance, render & extraction, and AI visibility. Each check returns a pass/fail with a weight and a prioritised fix recommendation.
Each category score = earned weight / total weight × 100. The overall score uses the same calculation across all categories combined. Fixes are sorted by weight so the highest-impact items appear first.
When enabled, the scanner auto-detects the site's language, generates relevant prompts, and fires them at live AI search engines via OpenRouter — Perplexity, Gemini, Claude, ChatGPT, and more. This runs in the background so the page stays responsive. Results include citation rate per model, a per-prompt breakdown, competitor citation comparison, and an AI-written management summary.
Scan results are saved to your account history. You can export as PDF or JSON, and ask follow-up questions about your results via an AI-powered chat that has full context of your scan.
The scanner is pointed at its own URL during development to verify it practices what it preaches. Every improvement made based on the scan results is committed to the codebase.
Scan this site now