Free tools
Both run in a second, ask for nothing but a URL, and give you the evidence rather than a mystery score. They cost us one HTTP request each — which is exactly why they can stay free.
robots.txt + llms.txt, 26 crawlers checked
Two files decide what an AI engine can do with your site: robots.txt says who may come in, llms.txt says what is worth reading. We split the crawlers into the ones that power answer engines — and can cite you — and the ones that only take training data. Most tools lump them together and call any block a problem. It isn't.
Two requests: /robots.txt and /llms.txt
Can an answer engine actually use this page?
Paste a URL and we run the page-level half of our real check registry against it — structure, schema, answer readiness, citability. Same engine as the product, narrower scope.
One request to the page
Why only two
Plenty of tools give away "AI visibility checks" that quietly query ChatGPT, Perplexity or Google AI Overviews on your behalf. Those queries cost real money per run, so somebody is paying — usually with heavy caching, aggressive throttling, or a number that was never live.
We would rather be straight about it. Anything that reads your own site is cheap and stays free. Anything that interrogates an answer engine costs us per run, so it sits inside the product where it can be budgeted and repeated on a schedule — which is the only way that number means anything anyway.