Can AI assistants actually read your site?
Most "AI SEO" advice is guesswork. This checks what is mechanically verifiable: whether ChatGPT, Perplexity, Claude and AI Overviews can reach your page, read it without running JavaScript, and quote it. Free, no signup.
See a real report: a live scan, not a mockup. Or read the aggregate data, or why we built this.
median of 191 pages scanned so far
What SEO doesn't cover
SEO is about getting the click. AI assistants don't send a click — they read the page, quote it, and answer without the reader ever visiting. Your site is either cited inside that answer, or it isn't there at all.
Existing tools don't measure this. Google Search Console tells you about Google. It says nothing about the crawlers that feed ChatGPT, Perplexity, Claude, or AI Overviews — and Cloudflare has blocked AI crawlers by default on every new site since September 2025, which means a lot of sites launched in the last year are already invisible to the assistants their owners want to be found by.
What gets checked
Whether AI search crawlers are blockedverifiable
Kept separate from training crawlers. Blocking GPTBot does not remove you from ChatGPT search. Blocking OAI-SearchBot does. Most tools conflate the two and give you the wrong advice.
Whether your server actually serves themverifiable
We request your page as an AI crawler. A CDN or WAF bot rule can return 403 while your robots.txt says yes. Invisible from the outside, and the most common silent failure.
Whether your content exists without JavaScriptverifiable
AI crawlers generally do not run JS. A client-rendered page looks empty to them.
And what we refuse to charge you forno evidence
We score llms.txt at zero and say so on the report. Google states it has no effect and almost none are ever fetched. Other tools sell it as a service. We won't.
Nothing breaks loudly
A scan is a photograph. What it photographs keeps moving, and so does the camera.
- A colleagueOne tick of a bot-protection box in your CDN. No deploy, no commit, nothing in your release notes.
- Your hostCloudflare has blocked AI crawlers by default on new sites since September 2025, so a migration can arrive already blocked.
- You, being carefulBlocking GPTBot stops training. OAI-SearchBot is the one that feeds ChatGPT’s answers, and most advice online treats them as the same thing.
- Your frameworkA rendering default moves to the client and your text stops existing in the HTML that crawlers read.
- Nobody at allNew assistants ship new crawlers. A clean bill of health last year was a clean bill of health for last year’s crawlers.
So the paid tier is not an alarm you hope never rings. It looks every week, tells you the day something moves, and keeps the record that shows it did not.
Pricing
Scan any page, as often as you like.
- Every check and every explanation
- No account, no card, no email
- A shareable report link
Watch one site. We look every week and write only when something moves.
- We ask your server for the page as an AI crawler, rather than taking robots.txt at its word
- A fixed sample of eight pages, so a template change shows up
- Severity-graded alerts, with retrieval-fatal breaks named as such
- Score history, so you can prove a fix held
Ten client sites. The same checks, one bill, one place to look.
- Everything in the single-site plan, across ten sites
- For people answerable for someone else’s visibility, not their own
- More than ten? Tell us how many and we will size the next tier
Prices exclude sales tax. Dodo Payments adds VAT or sales tax for your country at the checkout and shows the total before you pay, so the number you approve is the number you are charged. If your business is VAT-registered, enter your number there and it comes off.
We do not charge for llms.txt files, "AI optimisation scores" with no method behind them, or anything else we cannot show you the evidence for. Full pricing and what you are actually paying for →