See your page the way an AI crawler does. Pick GPTBot, Googlebot, PerplexityBot or ClaudeBot, enter a URL, and the checker requests the page with that crawler's user agent, then reports what came back: speed, format, title, description, structured data and a crawlability score with fixes.
Needs an account · 30 credits per scan
An AXP checker is a tool that requests a web page as an AI crawler would and shows what comes back. AXP means agent experience: what bots such as GPTBot, ClaudeBot and PerplexityBot receive from your site. This checker fetches the page with the crawler's user agent and reports the response time, format, title, description, structured data and a 0 to 100 crawlability score with fixes.
One request per scan, made with the user agent you choose, so you see the response that bot would get.
ChatGPT (GPTBot), Gemini (Googlebot), Perplexity (PerplexityBot), Claude (ClaudeBot), or a normal browser to compare against.
How long the page took to answer and whether it returned markdown or HTML.
The title and meta description the bot received, and every JSON-LD type found on the page.
A 0 to 100 score with findings, written by an AI analysis of the response.
Whether the site serves a crawler-specific version (an agent experience) to AI bots: configured, partly configured or not configured.
Advice for ChatGPT, Gemini, Perplexity and Claude, a list of action steps, and the raw content the bot saw, ready to copy.
Compare a crawler's view of your page with a browser's.
Pick the AI bot to simulate, or Chrome Browser to see the human version for comparison.
Paste the address of the page you want to test. The scheme is added if you leave it out.
Sign in and run it. Each scan uses 30 credits from your AI Rank Lab balance.
Check the score, the format the bot received, the schema found and the findings.
Work through the action steps, then scan again with a different crawler to compare.
Compare the AI crawler's result with the Chrome browser result for the same URL. The differences are what to fix.
0 to 100, from an AI analysis of the response. Read the findings under it rather than the number alone.
Most sites return HTML. Markdown means the site serves bots a lean text version, which is what AXP set-ups do.
Configured, partly configured or not configured: whether the site sends AI crawlers a version built for them.
What the bot received. If they are empty here but fine in a browser, the tags are probably added by JavaScript.
The structured data found in the response. Missing types that you see in a browser point to client-side rendering.
The text the crawler actually received. If your main answer isn't in it, the crawler can't use it.
Terwijl generatieve zoekmachine-bots zoals ChatGPT, Gemini en Claude websites scrapen, botsen ze vaak op conflicten tussen premium gebruikerservaringen voor mensen (visuele elementen, interactieve JS-componenten) en de vereisten voor ruwe citatie-indexering.
An AXP Checker (or agentic checker) audits whether your website returns crawler-optimized overrides. When a request matches an AI crawler user-agent (e.g. GPTBot, ClaudeBot, PerplexityBot), AXP intercepts it at the edge middleware level and serves a crawler-optimized Markdown page structure instead of generic browser HTML.
Many AI crawlers fetch the HTML your server sends and don't run JavaScript. A page that builds its content in the browser can reach them as a near-empty shell, even though it looks complete to a visitor.
Servers and CDNs can also treat bots differently: bot-protection rules, rate limits, geo-blocking or cookie walls may return an error or a challenge page to a crawler's user agent. Checking the page as each crawler is the quickest way to find out.
The fixes are the same basics that help search engines:
Something else? Ask the team
AXP stands for agent experience: serving AI crawlers a lean, crawler-friendly version of a page. The checker requests your page as a chosen AI bot and reports whether such a version is served, plus how readable the response is.
GPTBot for ChatGPT, Googlebot for Gemini, PerplexityBot for Perplexity and ClaudeBot for Claude, plus a normal Chrome browser so you can compare the two responses.
It needs a signed-in account and uses 30 credits per scan, because each scan runs a live request and an AI analysis. New accounts get 100 free credits with no card.
It shows what your server returns to that user agent. Sites that verify bots by IP address, or that block unknown traffic at a firewall, can still treat the real crawler differently.
Set crawl rules, publish an llms.txt file, and check what search engines see.
Validate an llms.txt file against the spec.
Create an llms.txt file for your site.
Build crawl rules for search and AI bots.
Test a live robots.txt for errors and blocked bots.
JSON-LD for articles, products, FAQs and more.
Score a page for search engines and AI answers.
100+ SEO, AI-search and developer tools.
The checker simulates one request. AI Rank Lab's AI Bots Tracker records the real visits from AI crawlers and AI assistants on your site.
Add one tracking tag and see visits from AI crawlers and people referred by AI assistants, page by page, so you know which pages AI engines read.
AI Bots Tracker
Also records visits referred by AI assistants, per page.
Paste your site. The free audit scores it for SEO, answer engines and generative engines, with the fixes.
