Skip to main content
Free tool · 10 seconds

Is my site visible to AI?

ChatGPT, Claude, Perplexity, Gemini… Enter your address and you’ll see which AI crawlers can read your site, and which lines to change.

We only read two public files: /robots.txt and /llms.txt. Nothing is stored.

Sources: each crawler’s official documentation
  • Google — Googlebot
  • Google — Google-Extended
  • OpenAI — OAI-SearchBot, GPTBot, ChatGPT-User
  • Anthropic — Claude-SearchBot, ClaudeBot, Claude-User
  • Perplexity — PerplexityBot, Perplexity-User
  • Apple — Applebot, Applebot-Extended
  • Meta — Meta-WebIndexer, Meta-ExternalAgent, Meta-ExternalFetcher
  • Mistral AI — MistralAI-Index, MistralAI-Training, MistralAI-User
  • Amazon — Amzn-SearchBot, Amazonbot, Amzn-User
  • Common Crawl — CCBot
  • How the rules are read: RFC 9309 (Robots Exclusion Protocol) · llms.txt: llmstxt.org
  • List checked on November 19, 2026. Only crawlers whose vendor publishes an official page are included.

Opening the door isn’t enough. Getting cited is what counts.

Leave your details: I’ll get back to you with an initial take on your visibility in AI answers (GEO audit). Your test result is attached automatically. Free, no commitment.

Email or phone — at least one of the two, so I can get back to you.

Your result is attached automatically. Your details are used only to get back to you — never sold, no newsletter.

Understanding the result

Three kinds of crawlers, three different decisions

AI vendors no longer rely on a single crawler. Most publish several, each with a documented role, and that’s what lets you decide with precision:

  • Search crawlers (OAI-SearchBot, Claude-SearchBot, PerplexityBot, MistralAI-Index, Meta-WebIndexer…) index your pages to cite and link to them in answers. Block them and you disappear from those answers. For Google, AI Overviews and AI Mode rely on Googlebot, according to Google.
  • Training crawlers (GPTBot, ClaudeBot, MistralAI-Training, Meta-ExternalAgent, Applebot-Extended, Google-Extended…) collect text to train future models. Allowing them or not is an editorial choice: it doesn’t change your presence in answers that cite sources.
  • “On-demand” crawlers (ChatGPT-User, Claude-User, Perplexity-User…) read a specific page when a user asks for it. Several vendors warn that they may not follow robots.txt.

The most common trap isn’t a deliberate block: it’s a robots.txt copied from a template, a “*” rule that’s too broad, or bot protection that makes the file return an error. The first two take five minutes to fix; the third makes the site invisible without anyone noticing. To go further, read the complete GEO guide or see the AI search optimization service.

Frequently asked questions

What exactly does this tool check?
It reads two public files on your site: /robots.txt, which tells crawlers what they’re allowed to explore, and /llms.txt, a summary meant for AI. For each AI crawler documented by its vendor (OpenAI, Anthropic, Google, Perplexity, Apple, Meta, Mistral AI, Amazon, Common Crawl), it shows whether your home page and your inner pages are open to it, and where the rule comes from. It doesn’t crawl your pages and stores nothing.
Will an “allowed” crawler necessarily cite me?
No. Access is necessary, not sufficient. An AI assistant cites the pages it considers reliable and relevant: precise content, sources, an identifiable author, structured data, mentions of your brand elsewhere on the web. The tool checks the door; what happens once the door is open is a matter of GEO.
Can you be cited by AI without feeding its training?
Yes, with most vendors, who keep their crawlers separate: OAI-SearchBot (search) and GPTBot (training) at OpenAI, Claude-SearchBot and ClaudeBot at Anthropic, MistralAI-Index and MistralAI-Training at Mistral AI, Meta-WebIndexer and Meta-ExternalAgent at Meta. Blocking the second ones doesn’t stop the first. The tool generates the lines to add. One nuance on the Google side: Google-Extended covers both Gemini training and the grounding of its answers, with no effect on Google Search or AI Overviews, which depend on Googlebot.
Why can some crawlers “ignore robots.txt”?
OpenAI (ChatGPT-User), Perplexity (Perplexity-User), Meta (Meta-ExternalFetcher) and Amazon (Amzn-User) state in their documentation that their “on-demand” crawlers, which read a page because a user asked for it, may not follow robots.txt rules. For them, the file is therefore no guarantee of a block.
Is an llms.txt file essential?
No. It’s a proposal published in 2024 (llmstxt.org): a Markdown file that presents your site and its key pages to language models. None of the vendors listed says, in its crawler documentation, that it reads it. It’s cheap to create and can help tools or agents that look for it, but it comes after a correct robots.txt, solid content and structured data.
My robots.txt can’t be found: is that a problem?
No: without a robots.txt (404 error), every crawler is allowed. The serious case is the opposite: a robots.txt that returns a server error (5xx) or doesn’t respond. The RFC 9309 standard then tells crawlers to treat the whole site as off-limits — a common, silent cause of invisibility, often due to a firewall or overzealous bot protection.

Your site is open to AI. Is it getting cited?

Describe your needs in a few lines: I reply within 24 hours with an initial analysis and concrete recommendations.