Free Tool

AI Citeability Checker

See your page the way an AI answer engine sees it.

Paste a URL. This fetches the page the way a crawler does, without running any JavaScript, and reports what an assistant can read, understand, and quote back to someone asking about your business. Results appear on this page. Nothing is stored.

One page at a time, over https. The check reads the page, your robots.txt, and your llms.txt. It takes a few seconds.

What It Checks

Thirteen checks, three questions

Being cited by an assistant needs three things to be true in order. The page has to be fetchable, then legible, then worth quoting. Each check below belongs to one of those.

Can it read the page at all

  • Content in the raw HTML

    The crawlers behind AI answers largely do not run JavaScript. A page that assembles itself in the browser arrives empty, and an empty page cannot be quoted.

  • AI crawler access

    Your robots.txt decides which crawlers may fetch the page. The check separates the ones that train models from the ones that fetch a page to answer a question someone is asking right now, because blocking those two has very different consequences.

  • Snippet directives

    noindex keeps the page out of the index. nosnippet allows indexing but blocks the extract, so the page can rank and still never be quoted.

Can it tell what the page is

  • Structured data

    JSON-LD states what the page is rather than leaving it to be inferred. Missing or malformed markup throws that away.

  • Semantic landmarks

    A <main> element separates the content from the navigation and footer. Without it, a parser has to guess which text is the page.

  • Title and meta description

    These become the label and the summary wherever the page is cited. Too long truncates, too short wastes the space.

  • One H1, headings in order

    The heading outline is how a machine reads the shape of an argument. Skipped levels break it.

Is there something worth quoting

  • An answer near the top

    A page that states its answer before the first subheading gives an assistant a passage it can lift. A page that builds to the answer over 800 words does not.

  • Canonical URL

    Without one, variants of the same page compete with each other and split whatever authority the page earns.

  • A publish or update date

    An undated page cannot be preferred over an older answer, because nothing says it is newer.

  • llms.txt

    Reported, not scored. Support for the convention is still uneven and its absence is not a defect.

The Limits

What this cannot tell you

An automated check is worth exactly as much as its honesty about what it missed. Four things are outside what any single-page scan can see.

Whether anything actually cites you

This reads one page and reports what it finds there. It does not query ChatGPT, Claude, Perplexity, or Gemini, and it cannot tell you how often any of them mention your business. That question needs a different method, which Blog #17 covers.

What the page looks like once JavaScript runs

The check deliberately stops where a non-rendering crawler stops. If your page fails the raw-HTML test and looks fine in a browser, both readings are correct, and the gap between them is the finding.

Whether your content is any good

Every check here is mechanical. A page can pass all thirteen and still say nothing worth quoting, and no automated check will ever tell you which one you have.

The rest of your site

One URL, one report. Site-wide problems, competing pages, and the internal linking that decides which of your pages gets surfaced at all are invisible from a single page.

The same limit applies to every automated scanner, including the ones that do not say so. We hold accessibility scanning to the same standard: an automated pass catches roughly 30% of WCAG issues, and we wrote up exactly what it misses rather than leaving it as a footnote.

FAQ

Frequently Asked Questions

What does citeable by AI actually mean?
It means an assistant answering a question can fetch your page, work out what it says, and quote a passage back with a link. That is a different goal from ranking. A page can sit high in search results and still be unquotable, usually because the text only exists after JavaScript runs or because a directive blocks the snippet.
Why does the check ignore JavaScript?
Because most of the crawlers feeding AI answers do the same. Rendering every page would report what a browser sees, which is the more flattering number and the less useful one. Stopping where a crawler stops is the point of the check.
Does this tell me how much traffic I get from AI assistants?
No. Referrals from assistants are largely invisible in analytics, and no page-level scan can recover them. This check answers the upstream question of whether your page is eligible to be cited at all.
Do you store the URLs people check?
No. The check runs, returns findings to your browser, and keeps nothing. There is no account, no history, and no record of what was checked.
Should I block AI crawlers in robots.txt?
It depends which ones. Training crawlers collect content to train future models, and blocking those is a defensible licensing decision with no effect on today's answers. Answer-time crawlers fetch a page because someone is asking a question now. Blocking those removes you from the answer. The check reports the two separately for that reason.
The check passed. Am I done?
It means nothing mechanical is in the way. Whether your page is the one an assistant picks depends on whether it answers the question better than the alternatives, which is an editorial question rather than a technical one.

Findings you would rather not fix yourself

Most of what this check reports is a half-day of work for whoever owns the codebase. When the answer is that the content only exists after JavaScript runs, it is an architecture problem instead, and that is the kind of thing we get called about. Send us the URL and what the check said.

Talk to us about the findings

How to Make Your Site Citeable by AI

The method behind these checks, and what to do about each one.

Do You Still Need a Website in the AI Search Era?

Why citeability replaced ranking as the thing worth optimizing for.

How to Measure AI Search Traffic

The question this tool cannot answer, and the triangulation that can.