Can ChatGPT read your website?
AI assistants such as ChatGPT, Perplexity and Copilot can only quote from your website what their crawlers are allowed to fetch and able to read. Sometimes robots.txt, a firewall or the page itself shuts them out without anyone having decided so. Enter your address: in half a minute you will see whether that is the case for you.
- Enter your address, no sign-up
- Free result: what works, what is missing
- Full report as a PDF,
€149Free until 30 Nov 2026 · then €149 one-time, no subscription
Checked:
Your badge is ready
This website passed every check. Paste the code into your website; the badge links to the public proof page and expires visibly after 90 days without a new check.
Your company block, written for you. Free.
Built from your own home page. Paste it into the head of your home page, and AI systems and search engines read your name, website address and description in machine-readable form.
What your site would look like to AI systems
First comes what your home page shows today, then the corrected version, built only from what your own page already says.
€149 Free until 30 Nov 2026 · then €149 one-time, no subscription
The full report as PDF for your AI Access Check: free until 30 November 2026, no card. We send you a confirmation email; one click in it opens your report. One report per business.
From 1 December 2026 a report costs €149 one-time again.
If you already pay for AI tools or an agency, the AI Reality Check shows in five days what actually delivers. Your plan states the price before anything starts.
If you would rather not brief your own agency or webmaster, have us implement it.
What the result looks like
What is checked
We read your robots.txt by the same rules the crawlers use, fetch your home page under the name of each crawler and compare the answer with the one a browser gets. In addition: noindex, text without JavaScript, title and description, company details as schema.org, phone and email, sitemap.
| Crawler | Company | Used for |
|---|---|---|
| OAI-SearchBot | OpenAI | Search in ChatGPT |
| ChatGPT-User | OpenAI | Fetch when a user asks ChatGPT about it |
| GPTBot | OpenAI | Model training |
| PerplexityBot | Perplexity | Search in Perplexity |
| ClaudeBot | Anthropic | Model training |
| Claude-SearchBot | Anthropic | Search in Claude |
| bingbot | Microsoft | Bing search and Copilot |
| Googlebot | Google search and AI Overviews | |
| Google-Extended | Use for Gemini | |
| Applebot-Extended | Apple | Use for Apple Intelligence |
| CCBot | Common Crawl | Open archive, the basis of many models |
What the check does not do
It does not promise that an assistant will recommend you; it checks the precondition for that. It does not tell you whether to admit training crawlers; that is your decision. It changes nothing on your website and gives no legal assessment.
- How accurate is it?
- We fetch under the crawlers' names, not from their addresses. A firewall that recognises crawlers by IP address may answer the real crawler differently.
- What happens to my address?
- It is used for the check and deleted with the result after 24 hours. A website that passes every check gets its badge automatically; its host, check date and score are then public.