Free AI visibility scan · no account · about twenty seconds
The free AI visibility scan, and what it can tell you
Give it a public web address. It reads what an answer engine would read and returns a sheet. It is the part of the audit a machine can finish on its own.
What the scan does
It fetches your robots file and sitemap, follows what they declare, and reads a bounded sample of pages as raw HTML, before any script runs, the way a retrieval system would. Then it checks whether pages return successfully and consistently, whether crawler directives look deliberate, whether a sitemap is present and coherent, whether canonicals are declared, whether structured data parses and what it declares, whether headings are stable and linkable, and whether body text is real text or unrendered template output.
Each of those has a yes or no answer that needs no judgement, which is why the scan can be free and fast.
What the scan cannot tell you
Whether a passage answers the question it sits under, whether a claim is material, whether a credential means what it appears to, whether comparison content is fair, or whether the named author knows the subject. Those need a person to read the page.
So the scan gives no score. It reads back what your site declares about itself and shows where the machine stopped. In the paid audit, a person reads the evidence for the rest and writes down a reason for each rating. Neither tells you what an engine said about you; that is a different product category.
What it does to your site
It requests public pages and nothing else. It names itself in its user-agent, respects your robots directives, stays inside published bounds on pages, depth, response size, redirects and time, and pauses between requests. It never signs in, submits a form, or makes a request that changes anything.
It refuses destinations that are not public web addresses (loopback, private and link-local ranges, cloud metadata endpoints, and redirects into any of them), so nobody can point it at an internal network. The full bounds are in the terms.
What happens to the result
The sheet and the scanned address are stored so the link keeps working. We do not store your IP address; rate limiting uses a truncated hash of it. No account, no email required, no follow-up sequence. The privacy page lists what is kept.
Going further
There is a real scan sheet published as it came out, and two complete paid audits: the operator's own site, unedited with its faults and a correction notice, and a redacted audit of a site we do not own. If you are weighing this against an AI visibility monitoring subscription, the two answer different questions, and the order you buy them in matters.