# Free AI visibility checker > This free scan checks whether AI crawlers can reach and read up to 30 public pages of your site, and quotes the evidence for every problem it finds. It does not ask ChatGPT or any other engine about your brand, so it measures technical readiness, not whether you are mentioned. No account, no card. Source: https://citetune.com/scan ## Two things “AI visibility” can mean One is whether AI engines mention or cite you. That is measured by asking real questions on each engine, repeatedly, and counting what comes back. The other is whether their crawlers can reach and read your pages at all. That is measured by fetching your site. The free scan does the second. The second comes first in practice: a page that a bot cannot fetch, or that carries no text in its raw HTML, cannot be cited however good it is. Sampling AI answers is part of the paid plans, where each surface reports its completion rate and an unread surface shows as not measured. ## How the scan runs It identifies itself as our crawler, obeys robots.txt, reads robots.txt and your sitemap, fetches up to 30 public pages and then applies the rule set. That set has sixteen active rules, covering robots.txt, the status codes bots receive, noindex and canonical tags, redirects, sitemaps, hreflang, Organization JSON-LD and page evidence. No AI model is called. Progress follows real page counts, never a fake spinner. If we are rate-limited we show the result we already have instead of an error. ## FAQ ### Does the free scan show whether ChatGPT mentions my brand? No. The free scan is technical only. It fetches your pages and quotes what it observed; it does not ask ChatGPT, Gemini or any other engine about your brand, so it cannot say whether you are mentioned or cited. Sampling real AI answers is part of the paid plans. ### Does the free scan run JavaScript? No. In production it reads the raw HTML a crawler receives and does not render pages in a browser. Text that only appears after a script runs is therefore not seen by the scan, which is worth knowing if your site is built with React or another client-side framework. ### How many pages and how often can I scan? Up to 30 public pages per scan, within about 90 seconds, and one scan per domain per day: scanning the same domain again that day shows the existing result. Results are deleted after seven days unless you claim the report with a free account. ### Does the scan respect my robots.txt? Yes. It identifies itself as our crawler, obeys robots.txt and records which AI bots your robots.txt allows or blocks. If your site refuses most requests with a 403 or 429, the scan says so and reports the limited evidence it collected instead of guessing. ## What the free scan does - Fetches up to 30 public pages of your site as an identified crawler, and obeys robots.txt. - Reads robots.txt and your sitemap, and records which AI bots robots.txt allows or blocks. - Runs deterministic rules over what it fetched and quotes the observed response (status codes, headers, HTML excerpts) behind each finding. - Shows the first finding in full, with a proposed fix, its safety tier and a rollback. Claiming the report with a free account unlocks the next two findings. - Needs no account and no card, finishes in about a minute, and deletes the result after seven days. ## What it does not do - It does not run JavaScript. In production the scan reads the raw HTML a crawler receives, so text that only appears after a script runs is not rendered or checked. - It does not ask ChatGPT, Gemini or any other AI engine about your brand. The free scan is technical only: no AI answers are sampled, so it cannot tell you whether you are mentioned or cited. - It does not give a score or a grade. It lists facts and findings. - It does not change your site. Fixes are patches you review and apply. - It does not cover more than 30 pages, or the same domain more than once a day. Limits in production: 30 pages, 90 seconds, one scan per domain per day. Measuring AI answers is part of the paid plans, which show the method, completion rate and unread surfaces for each. ## What a finding looks like This is the engine’s own output for rule site_blocks_crawlers on our reference configuration, not a customer’s site. Observed facts come first, quoted; the mechanism we infer from them is kept apart, with the alternative explanations that would change it. Rule: site_blocks_crawlers **Observed facts:** 38 of 42 fetches were refused with 403 (status codes: 403 x38, 200 x4). **Evidence excerpt:** ``` GET / HTTP/1.1 user-agent: HTTP/1.1 403 Forbidden ``` **Inferred mechanism:** Your site returns 403 to non-browser clients — AI crawlers likely see the same. An anti-bot layer (WAF/CDN bot management) is rejecting requests before they reach your pages, so GPTBot, ClaudeBot, PerplexityBot and similar agents cannot read your content at all. **How to verify:** curl -A "GPTBot" https:/// and confirm a 200 with the page HTML instead of a 403. ## Read next - [How we measure, and what we call not measured](https://citetune.com/methodology) - [One finding, from evidence to a verified fix](https://citetune.com/how-it-works) - [AI bot access checker](https://citetune.com/tools/ai-bot-access-checker) - [Why ChatGPT does not cite your site](https://citetune.com/blog/why-chatgpt-doesnt-cite-your-site) - [Pricing](https://citetune.com/pricing)