AI visibility score: can AI find your site?
See whether ChatGPT, Claude, Perplexity and Google can find, read and quote a page. Enter a URL and get a score out of 100, which crawlers are allowed in, and a fix for every gap.
- Free, no signup
- Scored in your browser
- Nothing is stored
The address goes to my checker, which reads the page, robots.txt, llms.txt and sitemap once, the way an AI crawler does, and keeps nothing. The score is worked out in your browser.
How it works
How to use the AI visibility score
- 01
Enter a page
Your homepage, or the page you most want AI tools to quote: a pricing page, a service page, a guide.
- 02
Let it read the page
The checker fetches the raw HTML once, without running JavaScript, plus robots.txt, llms.txt and the sitemap, the way an AI crawler does.
- 03
Read the score
19 checks in five groups add up to 100: crawler access, readable HTML, structured data, trust signals and how easy the page is to quote.
- 04
Fix the gaps
Every check that misses says what to change. Start with crawler access and readable HTML, which matter most.
How the score works
The score has 19 checks in five groups, weighted by how much each one decides whether an AI tool can use the page at all. Access and readable HTML come first, because nothing else matters if a crawler is turned away or sees an empty page.
| Group | Points | What it checks |
|---|---|---|
| Crawler access | 30 | robots.txt, eight AI search and answer crawlers, noindex, and a sitemap |
| Readable without JavaScript | 25 | Text in the raw HTML, one H1, section headings, a declared language |
| Structured data | 20 | JSON-LD, who is behind the site, what the page is, title and description |
| Trust signals | 15 | HTTPS, a canonical URL, an author or publisher, a published or updated date |
| Easy to quote | 10 | Question-style headings or FAQ markup, and an llms.txt |
80 or more is ready for AI. 50 to 79 is partly ready. Under 50 means AI tools will struggle to use the page.
Search crawlers and training crawlers are different
The AI companies now run separate crawlers for separate jobs, and each one is controlled by its own line in robots.txt. Search and answer crawlers, such as OpenAI's OAI-SearchBot and ChatGPT-User, Anthropic's Claude-SearchBot and Claude-User, and Perplexity's PerplexityBot and Perplexity-User, fetch pages so they can appear in answers, often with a link. Training crawlers, such as GPTBot, ClaudeBot, Google-Extended and Applebot-Extended, collect content that may be used to train models.
This tool scores only the first kind. Blocking training crawlers is a reasonable business decision and costs you nothing here. Blocking the search crawlers is usually an accident, often a blanket rule copied from a template, and it quietly keeps you out of AI answers.
The JavaScript problem
Many modern sites send a near-empty HTML file and build the page with JavaScript in the browser. Google renders that JavaScript. According to Vercel and MERJ's December 2024 study, the major AI crawlers it tested from OpenAI, Anthropic and Perplexity do not. So the "content in the HTML" check is worth 10 points on its own: it is the most common reason a good page is invisible to AI tools. The fix is server-side rendering or static generation, which every serious framework supports.
What the score cannot see
- Firewalls. Hosting providers and CDNs can block crawlers before they reach robots.txt. If your host has a setting that blocks AI bots, it can override everything here.
- Whether you are recommended. That depends on your content and on how often other sites cite you. To see who AI tools actually recommend for your searches, see AI visibility tracking.
- The rest of the site. The score covers one page. Check your homepage, your pricing page and your most important guide separately.
Fixing it
Most fixes are small: a robots.txt rule, a block of structured data, a few headings phrased as questions. Write an llms.txt with the llms.txt generator and structured data with the schema generator. If you would rather have it done across the whole site and kept up every month, that is the SEO + GEO plan.
Sources
Each crawler's job was checked in its company's own documentation on 27 October 2026:
- OpenAI: overview of OpenAI crawlers
- Anthropic: does Anthropic crawl data from the web
- Perplexity: Perplexity crawlers
- Google: Google's common crawlers, including Google-Extended
- Apple: about Applebot
- Vercel and MERJ: the rise of the AI crawler
- RFC 9309: the robots exclusion protocol, which the robots.txt check follows
- llmstxt.org: the llms.txt proposal
Common questions
AI visibility score: FAQ
Whether AI tools can find, read, understand and quote a page. It checks that AI search crawlers are allowed in, that the content is in the HTML rather than added by JavaScript, that structured data says who you are and what the page is, that the page shows trust signals such as an author and a date, and that it answers questions directly.
No. The score measures whether AI tools can use your page, not whether they will choose it. Being recommended also depends on what the page says, how well it answers the question, and how often other sites mention you. A low score, though, can rule you out before any of that counts.
Block the training crawlers if you do not want your content used to train models: GPTBot, ClaudeBot, Google-Extended and Applebot-Extended. Allow the search and answer crawlers, such as OAI-SearchBot, Claude-SearchBot and PerplexityBot, if you want to appear in AI answers. The companies document these as separate controls, so you can do both.
Because that is what most AI crawlers see. A December 2024 study by Vercel and MERJ found that the major AI crawlers it tested from OpenAI, Anthropic and Perplexity do not run JavaScript, while Google's and Apple's do. If your text only appears after scripts run, those crawlers see an almost empty page.
No. Google's documentation says Google-Extended controls whether content is used for Gemini training and grounding, and does not affect inclusion or ranking in Google Search.
Because it is a proposal, not a standard, and AI companies have not confirmed that their crawlers use it. It is cheap to add and can help when an AI tool or agent looks for it, so it earns 5 points out of 100, not more.
It goes to a small checker on this site, which fetches that page, robots.txt, llms.txt and the sitemap once and returns them to your browser, where the score is worked out. Nothing is stored and the URL is not logged. Only public websites can be checked.
Usually a firewall or bot protection answered with a challenge instead of the page. That is worth knowing: if it stops this checker, it may stop AI crawlers too. Other causes are a page that takes more than 8 seconds to answer, or more than 5 redirects.
Need it done for you?