How it works
- We request the page once from our server, following redirects, and keep the HTML exactly as it arrives.
- We don’t run any JavaScript. Vercel found that ChatGPT and Claude’s crawlers download JavaScript files but don’t execute them, so this is the version they read.
- We run the per-page checks from the GEOHATS Site optimizer and turn each issue into a fix, most important first.
What it checks
| Check | What we look for | Why it matters |
|---|---|---|
| Text in the raw HTML | Word count without running JavaScript; flags client-rendered pages | Most AI crawlers don't run JavaScript |
| noindex | Meta robots tag and X-Robots-Tag header | noindex pages stay out of the search indexes AI search draws on |
| Title | Present, 20-65 characters | The first signal of what the page is about |
| Meta description | Present, 70-160 characters | Engines show and reuse it instead of guessing |
| H1 | Exactly one | States the page's main topic |
| Answer-first block | First real paragraph is 80 words or fewer | Short, self-contained answers are easy to quote |
| Table or FAQ | An HTML table, FAQ heading, FAQPage schema or question headings | Formats AI lifts into answers |
| Structured data | JSON-LD types found on the page | Tells engines what kind of page it is |
| Date | dateModified or datePublished in schema, meta tags or a <time> element | Assistants look for fresh sources |
| Canonical | Present and pointing to this page | Decides which URL gets indexed |
| Long paragraphs | Paragraphs over 150 words | Walls of text are hard to quote |
| Image alt text | Images with no alt attribute (alt="" is fine) | Crawlers understand images through alt text |
| Links | Internal and external link counts | Internal links connect your related pages |
| Response time | Time to the full HTML | Live fetches give up on slow pages |
This tool checks one page. Site-wide access (robots.txt rules per AI crawler, firewall blocks, sitemap, llms.txt) is covered by the AI crawler check.
FAQ
What does "AI-ready" mean for a page?
Two things. AI crawlers can read the page (its text is in the HTML, it isn't marked noindex, it answers fast), and the page is easy to quote: a short direct answer near the top, a table or FAQ, structured data, a visible date and short paragraphs.
How is the score calculated?
Each issue costs points by severity: 25 for critical, 10 for a warning, 3 for a notice, from a start of 100. These are the same checks and weights the GEOHATS Site optimizer uses for one page. The answer-ready checks only count on pages with 600 words or more, since a short utility page doesn't need an FAQ.
Why does it say my page loads with JavaScript?
The raw HTML had almost no text and signs that the content arrives through JavaScript, such as an empty app container. ChatGPT, Claude and Perplexity's crawlers don't run JavaScript, so they get that empty version. Server-side rendering or static export fixes it.
Are empty alt attributes a problem?
No. alt="" is the correct way to mark a decorative image. This check only counts images that have no alt attribute at all.
Does a perfect score mean ChatGPT will cite my page?
No. A high score means the page is easy to read and quote. Whether an assistant picks it depends on how well it answers the question compared to other pages, and on whether other sites mention you.
Is it free? Do you store my page?
It's free and doesn't use any AI model. We fetch the page once, run the checks and show the result. Nothing is saved and you don't need an account.
Learn more
- The content AI assistants actually cite: page structure that wins
- You rank #1 on Google but ChatGPT ignores you: why, and the fix
More free tools
- AI crawler checkCan ChatGPT, Claude and Perplexity reach and read your site?
- robots.txt AI rules builderChoose which AI crawlers to allow and copy the lines.
- GA4 AI traffic regexTrack visits from ChatGPT, Perplexity and other assistants.
- Site optimizer (GEOHATS)Runs these checks on every page of your site and suggests the fixes.
By Daniel Borodin, GEOHATS founder · Last updated