Can AI read your website?
Find what stops AI reading your site — and exactly how to fix it.
ChatGPT
Claude
Gemini
PerplexityInstant preview · Full report free by email
55website checks completed
canairead.me
Example AI readiness report
canairead.me · 92/100
Built to test what actually determines AI visibility today
CanAIRead.me is an independent diagnostic tool. It is not affiliated with or endorsed by OpenAI, Anthropic, Google, Perplexity, xAI or Microsoft.
How it works
Three steps, no setup required.
From your website to a practical fix list.
Enter your website
Paste any URL. No account, no install, no code to add — we work from what's already public.
We analyse AI accessibility and readability
24 checks across crawler access, rendered HTML, structured data, entity clarity and citation signals.
Get your AI readiness report
A clear score, platform-by-platform breakdown, and specific fixes — ranked by what actually matters.
Five questions, not one score
“AI-friendly” isn't one thing. A site can be fully crawlable and still be impossible to classify, or perfectly structured and still blocked at the door. We test each independently.
Each category answers a different question.
Can AI systems reach the site?
We check robots.txt rules for each major AI crawler, response codes, and whether requests are blocked by a CDN or WAF before content is ever served.
Can useful content be extracted?
We compare the raw HTML response against the rendered page to see how much content depends on JavaScript, and whether headings and structure are machine-parseable.
Can the website be confidently classified?
We look for consistent, structured signals — title, meta description, headings, schema — that let a model determine who you are and what you offer.
Is useful factual information easy to identify?
We look for clearly structured facts, claims and Q&A content — the kind of content AI systems can lift and cite with confidence.
Are underlying search/indexing signals healthy?
Sitemap validity, canonical consistency and core indexability — the foundation most AI features, from Gemini to AI Overviews, are ultimately built on.
A note on llms.txt
We treat llms.txt as an experimental, emerging signal — useful to have, not a ranking or citation guarantee. It contributes a small part of one category, never the headline score.
Read the llms.txt guide →Passing these checks improves the conditions for AI access, understanding and citation. It does not guarantee that any specific AI system will recommend, rank or cite your website — no tool can promise that.
Know your crawlers
Every AI platform sends a different agent for a different purpose. Blocking the wrong one can quietly remove you from search, training or citations.
Search access and training permission are separate choices.
GPTBot
OpenAI
Crawls public content that may be used to train future OpenAI models.
Explore the related checker →OAI-SearchBot
OpenAI
Powers live web results and citations shown inside ChatGPT search.
Explore the related checker →ClaudeBot
Anthropic
Anthropic's general-purpose crawler for public web content.
Explore the related checker →Google-Extended
Controls whether your content can be used to improve Gemini and Vertex AI models, separately from search indexing.
Explore the related checker →PerplexityBot
Perplexity AI
Crawls and indexes content to power live citations inside Perplexity answers.
Explore the related checker →Claude-SearchBot
Anthropic
Powers live citations when Claude uses web search.
Explore the related checker →Questions worth asking
What does "AI can read your website" actually mean?
It means an AI system — ChatGPT, Claude, Gemini or Perplexity — can technically fetch your pages, extract meaningful content from the HTML, and correctly understand what your business does. Those are three separate things, and a site can pass one while failing another.
Is this the same as an SEO audit?
It overlaps, but it isn't the same. Traditional SEO audits focus on ranking in Google. We focus on whether AI crawlers can access your content, whether it's structured clearly enough to be understood, and whether it contains the kind of facts AI systems are comfortable citing.
Does a high score guarantee ChatGPT or Perplexity will cite me?
No — and we'd be skeptical of any tool that claims it does. A high score means the technical and structural conditions are in place. Whether a specific answer cites you still depends on relevance, competition and the query itself.
What's the difference between GPTBot and OAI-SearchBot?
GPTBot crawls content that may be used to train OpenAI's models. OAI-SearchBot is separate — it powers live citations inside ChatGPT search. You can allow one and block the other; they're controlled independently in robots.txt.
Do I need an llms.txt file?
Not urgently. llms.txt is an emerging, experimental convention with limited adoption so far. We flag it as a nice-to-have, not a priority — robots.txt, structured data and clean HTML matter far more today.
Is CanAIRead.me free?
Yes. Running a scan and seeing your score, platform breakdown and top findings is free with no account required. A deeper report with every finding is available by email.