Free AI crawler check
Give us a URL and we fetch it the way GPTBot, ClaudeBot and PerplexityBot do — no JavaScript, no cookies — then show you what actually came back, and what to change.
Free, no account. 10 checks a day, 30 with an account. We identify ourselves in the
User-Agent and respect robots.txt.
Visitors get the finished page. Most AI crawlers ask for the same URL without running JavaScript, so all they get is what is in the HTML. Everything we report comes from that second fetch.
GET /pricing · JavaScript not executed 200 OK · text/html <body> <div id="root"></div> <script src="/assets/app.8f2c1a.js"></script> </body> # nothing here a reader could quote
Illustration. Your report shows the HTML that actually came back, byte for byte.
25 checks in four dimensions, published in full. None of them predicts anything: each one looks for something that either is or is not in the response.
Can AI crawlers get in at all?
robots.txt rules for GPTBot, ClaudeBot, PerplexityBot and the rest — plus sitemap, redirects and status codes.
How much text is really there?
How much readable text sits in the HTML before JavaScript runs, and what only exists after it runs.
Does the page say who it is?
JSON-LD, entity declarations and language attributes: the parts a machine reads instead of guesses.
Are the fundamentals in place?
Title, meta description, canonical, headings, image alt text.
A finding you cannot check yourself is only an opinion. Each one names the rule, says why it matters and how to fix it, and attaches the material it judged — status codes, byte counts, the HTML fragment it came from.
EXTRACT-001 Content readability 312 characters 41,204 bytes An example, in the shape the report writes them. Your numbers are your own, and reports live at a permanent link you can send to a client.
The half of SEO that decides whether AI systems can use your pages.
What GPTBot, OAI-SearchBot, ClaudeBot and PerplexityBot can actually read.
Split retrieval crawlers from training crawlers, so blocking one doesn't cost you the other.
The small map of your site that AI systems can read.
The Organization and WebSite declarations that tell machines who you are.
The full list, sorted by what happens when you block each one.
Which file does what, and the mistake that costs people citations.
It doesn't give you a "GEO score" that predicts anything, and it won't claim to know whether a specific AI will cite you. It measures whether your content is readable by machines — a precondition, not a promise.
This site is checked by the same tool: 100/100, no findings. We run our own rules on ourselves before asking you to run them on your site. Every check is documented in public: see all the rules we run — what each one looks for, why it matters, and how to fix it.