Apex36 Labs
AI search readiness checker
By Yash Soni, Founder & CEO, Apex36Updated
In short
An AI search engine can cite a page only if its crawler is allowed in robots.txt, the text is in the HTML without running JavaScript, and the page states clearly what it is. This free checker tests one page for those things in 8 checks, in about 10 seconds, and returns a score out of 100 with fixes ranked by the points they add.
Key facts
| Cost | Free, no sign-up |
|---|---|
| What it checks | 8 checks on one page, scored out of 100 |
| What we fetch | The page, robots.txt, llms.txt and your sitemap. At most 4 requests, never a crawl. |
| Limits | 10 checks an hour per visitor; results cached for 24 hours |
Can ChatGPT, Perplexity and Google AI Overviews read my website?
They can if three things hold. Their crawlers, such as OAI-SearchBot, Claude-SearchBot and PerplexityBot, must be allowed in robots.txt; the words must be in the HTML the server sends; and the title, headings and structured data must say what the page is. Google states that AI Overviews and AI Mode need no special markup beyond being indexable in normal Search.
What does the checker test, and how is it scored?
It runs 8 checks on one page. Each passes (full weight), warns (half) or fails (none), and two checks give partial credit for each part that passes. The weights add up to 100, and crawler access and server-rendered content carry the most because without them nothing else is read.
| Check | Points |
|---|---|
| AI crawler access in robots.txtIf robots.txt blocks GPTBot, ClaudeBot or PerplexityBot, those engines cannot read the page, so they cannot cite it. | 20 |
| Content without JavaScriptMost AI crawlers do not run JavaScript. Text that only appears after scripts run is invisible to them. | 20 |
| llms.txtA short markdown file at /llms.txt tells AI tools what your site is and which pages matter. | 10 |
| Title, description, one H1 and canonicalThese tell every crawler what the page is about and which URL to cite. Partial credit for each of the four. | 15 |
| Structured data (JSON-LD)Schema names your company, services and page type in a form machines read without guessing. | 15 |
| Open Graph tagsog:title, og:description and og:image shape how a link looks when it is shared or shown as a source. | 5 |
| SitemapA sitemap listed in robots.txt helps crawlers find every page, not just the ones linked from your home page. | 10 |
| Page weight hints (estimate)Heavy HTML and blocking scripts slow crawlers and visitors. An estimate, not a Lighthouse score. | 5 |
Why do AI crawlers need content without JavaScript?
Because most of them do not run it. Vercel's analysis of AI crawler traffic, published 17 December 2024, found that ChatGPT's and Claude's crawlers fetch JavaScript files but never execute them, so text that appears only after scripts run is invisible to them. Google's Gemini is the exception, as it uses Googlebot's rendering.
What can't the score tell you?
The score measures whether AI engines can read the page, not whether they will choose it. Being cited also depends on how well the page answers the question and how far the wider web trusts the site, which no single-page check can measure.
- It checks one page, not the whole site. Other templates may pass or fail differently.
- The page weight line is an estimate from the HTML, not a Lighthouse or Core Web Vitals score.
- It reads robots.txt, not your firewall or CDN rules, which can block AI crawlers on their own.
- It does not see whether a page is indexed in Google or Bing.
Where do these rules come from?
Each check follows what the AI companies and Google publish about their own crawlers, plus one independent study of how those crawlers behave. The sources below were read on Nov 18, 2026.
- OpenAI, Overview of OpenAI crawlersOAI-SearchBot, ChatGPT-User and GPTBot, and what each is for
- Anthropic, How Anthropic's crawlers work and how to block themClaudeBot, Claude-SearchBot and Claude-User
- Perplexity, Perplexity crawlersPerplexityBot and Perplexity-User
- Google Search Central, AI features and your websiteHow AI Overviews and AI Mode pick pages
- Google Search Central, Google's common crawlersGoogle-Extended and what blocking it changes
- Vercel, The rise of the AI crawler (17 Dec 2024)Found that ChatGPT's and Claude's crawlers fetch JavaScript files but do not run them
- llmstxt.org, The /llms.txt proposalThe file format the llms.txt check looks for
What would Apex36 fix for you?
An AI SEO audit from $500 covers the whole site, not one page, with a ranked fix list. We then ship the fixes as code in your repo: AI crawler policy, server-side rendering, llms.txt, one schema graph and Core Web Vitals work in the templates.
Frequently asked questions
No tool can promise that. Passing means AI search engines can fetch, read and understand the page. Whether they cite it depends on whether it answers the question better than other pages, and on how far the wider web trusts your site.
Many AI crawlers read the raw HTML and do not run JavaScript. If your text only appears after scripts run in the browser, those crawlers see an empty page. Server-side rendering puts the text in the HTML from the start, so every crawler sees what a visitor sees.
Only if you do not want to appear in AI answers. Search crawlers such as OAI-SearchBot, Claude-SearchBot and PerplexityBot fetch pages so the engine can cite them. Training crawlers such as GPTBot, ClaudeBot and CCBot collect data for future models. You can allow search crawlers and block training ones; this checker counts GPTBot and ClaudeBot as major crawlers, so expect a lower score if you block them on purpose.
A markdown file at the root of a site that gives AI tools a short summary of the site and links to its most useful pages. Support among AI companies is still uneven, so it carries 10 of the 100 points here. It takes an hour to add.
No. Google says there are no extra requirements to appear in AI Overviews or AI Mode: the page must be indexed and eligible to show a snippet in normal Search. Clear structure and structured data that matches the visible text still help every engine understand the page.
Results are cached for 24 hours so repeat checks are fast; we keep a hashed form of your IP for rate limits for up to 7 days. If you ask us to email the report, we send it to you and a copy to our team; we don't add your email to a list or our database.