AI visibility check: Can ChatGPT & Co. read your website?
In a few seconds we check whether ChatGPT, Perplexity, Claude and Google AI Overviews are allowed to fetch your website, can understand its content and can cite you as a source. With a score, traffic-light rating and concrete recommendations. Free and without sign-up.
What we would tackle first
The check shows whether AI systems can read you. A GEO analysis shows whether they recommend you.
We put the questions your customers ask to ChatGPT, Perplexity, Gemini and Google AI and evaluate which providers are named, with which sources and why. The result is a plan for content, technology and reputation. Send us your result and we will get back to you with an honest assessment.
Twelve points in three areas: access, readability, identity.
AI systems can only recommend what they are allowed to read, are able to read and can clearly attribute. That is exactly what the check tests, using the same rules the crawlers themselves follow.
AI search crawlers in robots.txt
Whether Googlebot, Bingbot, OAI-SearchBot (ChatGPT), Claude-SearchBot and PerplexityBot may fetch your home page and subpages, evaluated according to the official standard RFC 9309.
Firewall and bot protection
Whether your server or CDN, Cloudflare for example, turns AI crawlers away even though robots.txt allows them. To find out, we fetch your home page the way the crawlers identify themselves.
Indexing and citability
Whether noindex, nosnippet or max-snippet:0 prevent Google and other systems from showing your content or quoting it in AI Overviews.
AI training
Which training crawlers you block, such as GPTBot, ClaudeBot or Google-Extended. As a note, not an error: that is your decision.
Readability without JavaScript
How much text is in the delivered HTML. Most AI crawlers do not run JavaScript and only see what the server delivers.
Title, description, headings
Page title, meta description, one clear H1, subheadings, language attribute and preview data.
Structured data
Whether Schema.org data describes your company in machine-readable form, with name, logo, address and profiles, and whether it is in the HTML or only injected via Tag Manager.
Answer-ready content
Whether there are questions with direct answers, for example in an FAQ or accordions, plus lists and tables. AI systems prefer to cite passages like these.
Response time
How quickly the home page is delivered. Crawlers work with tight time limits.
Identity and trust
Whether an imprint, contact details, an about page and profiles on other platforms can be found, so AI systems know who is behind the site.
XML sitemap
Whether a sitemap exists and is listed in robots.txt, including modification dates.
llms.txt
Whether an llms.txt summarises your company for language models. As a note only, because its benefit is not yet proven.
Not every AI crawler is the same: which ones to allow.
Many websites block “the AI” across the board and disappear from ChatGPT, Perplexity and the like as a result. Yet there are two kinds of crawler. Search crawlers fetch content to answer questions and link to sources. Block them and you will not be cited. Training crawlers collect content for training future models. Blocking them is a business decision that costs no visibility in AI search.
| Token | Operator | Purpose | Recommendation |
|---|---|---|---|
Googlebot | Google Search, AI Overviews, AI Mode | Allow | |
Bingbot | Microsoft | Bing and Microsoft Copilot | Allow |
OAI-SearchBot | OpenAI | Search in ChatGPT | Allow |
ChatGPT-User | OpenAI | Fetches when users ask ChatGPT to open a page | Allow |
Claude-SearchBot, Claude-User | Anthropic | Web search and page fetches in Claude | Allow |
PerplexityBot | Perplexity | Answers with cited sources | Allow |
GPTBot | OpenAI | Model training | Your decision |
ClaudeBot | Anthropic | Model training | Your decision |
Google-Extended | Gemini training and answers in Gemini, not Google Search | Your decision | |
CCBot | Common Crawl | Open web dataset, the basis of many models | Your decision |
Providers change their crawlers from time to time. The check is updated continuously.
What the check can and cannot do.
The check reliably shows whether AI systems can fetch, read and understand your site. That is the prerequisite for any visibility in ChatGPT, Perplexity, Claude and Google AI. Without it, even the best content will not help.
It does not show whether AI systems actually name your company. That also depends on how well your content answers your audience’s questions and how often other sources write about you. We examine that in a GEO analysis.
The home page is evaluated. Some servers turn away automated requests or show a security check; in that case we would rather show no result than a wrong one. And because real crawlers come from their own IP addresses, the firewall test is a strong indication, but not proof.
Questions about the AI visibility check
Is the AI visibility check free?
Yes. The check is free, with no sign-up and no obligation. You enter a domain and see the result, with score and recommendations, straight away in your browser.
What is GEO (Generative Engine Optimisation)?
Generative Engine Optimisation (GEO) covers everything that makes AI systems such as ChatGPT, Perplexity, Claude, Gemini or Google AI Overviews use a website as a source and name the company in their answers. The foundation is technical: the crawlers must be allowed to fetch the site and be able to read the content without JavaScript. Clear content, structured data and a recognisable reputation build on that.
Should I block AI crawlers in robots.txt?
It depends on the crawler. Search crawlers such as OAI-SearchBot (ChatGPT), Claude-SearchBot, PerplexityBot, Googlebot and Bingbot should have access; otherwise your site can neither be cited nor linked in AI answers. Training crawlers such as GPTBot, ClaudeBot, Google-Extended or CCBot collect content for training language models. Blocking them is a legitimate decision and costs no visibility in AI search. There is an overview in the table above.
Why is content loaded via JavaScript a problem?
The crawlers of ChatGPT, Claude and Perplexity generally do not run JavaScript. They only read the HTML the server delivers. If a page’s content is only built in the browser, as with some React or Vue applications, these crawlers see an almost empty page. Server-side rendering or static pre-rendering fixes this.
What is an llms.txt and do I need one?
An llms.txt is a file in the root of a website that summarises the company and its most important pages in Markdown for language models. It is a proposal from 2024 that the major providers do not yet officially use. The effort is small and the effect currently uncertain, so the check does not include it in the score.
Does this also apply to Google AI Overviews and AI Mode?
Yes. For AI Overviews and AI Mode, Google uses the regular Googlebot and the Google index. What matters is that the page is indexable and allows text snippets. With nosnippet or max-snippet:0 you exclude content from AI Overviews. The Google-Extended token, by contrast, only controls training and answers in Gemini, not Google Search.
Does the check show whether ChatGPT recommends my company?
No. The check tests the prerequisites: whether AI systems can fetch, read and understand your site. Whether they actually name your company for the relevant questions also depends on content and reputation. We examine that in a GEO analysis with real questions put to the AI systems.
What data do you store?
We only store the domain checked, the time, the score and the page language, no IP address and nothing about you. Details are in our privacy policy.
What is the difference between the check and a GEO analysis?
The check tests the technical prerequisites automatically, from the outside. In a GEO analysis we put the questions your customers ask to the AI systems, evaluate who is named and why, and derive concrete measures for content, technology and reputation.
Want to know whether AI systems recommend you?
We show you how ChatGPT, Perplexity and Google AI answer questions about your industry today, where you are missing and what you can do about it. And if you want to check your tracking at the same time: the free tracking check shows whether GA4, Consent Mode v2 and the Meta Pixel are set up properly.
Arrange a call