What is an AI readiness checker?
An AI readiness checker is a quick checkup for one of your web pages.
It answers a simple question: can AI tools find your page, read it, and quote it?
Tools like ChatGPT, Perplexity, and Google's AI Overviews no longer just list ten blue links. They write an answer and cite the pages they used. Being one of those cited pages is the new front page.
But AI engines do not work like the old Google crawler. They use their own bots, they often skip JavaScript, and they lean on clear signals they can trust. So a page that ranks fine today can still be invisible to AI. This checker tests for that, one URL at a time, in a few seconds.
You will sometimes see this called GEO, short for generative engine optimization. It is the same idea as SEO, just pointed at AI answers instead of only search rankings.
The two things that decide AI readiness
Most of AI visibility comes down to two questions, in this order.
Can the bot reach the page? This is AI crawler access. If your robots.txt blocks the bot, nothing else matters. You are simply not in the running.
Can the bot read the page? Once it is in, the bot still has to see your words and understand them. That means content in the raw HTML, clear structure, and signals like an llms.txt file or schema that tell the engine what the page is.
The checker scores both, then ranks the gaps so you know what to fix first.
| Readiness signal | Aim for | Why it matters |
|---|---|---|
| AI crawler access | Retrieval bots allowed in robots.txt | If the bot can't fetch the page, nothing else counts |
| Readable without JavaScript | Main text in the raw HTML | Many AI crawlers skip JavaScript and read source only |
| llms.txt | Present at the site root | Points AI models to your cleanest, best content |
| Structured data | Article / FAQPage JSON-LD | Lets an engine label and trust what the page is about |
| Indexable | No stray noindex, in the sitemap | An unindexable page is invisible to search and AI alike |
AI crawler access: can the bots even reach you?
This is the first and most important check.
Every AI company sends out its own bot to read the web, and each one has its own name. A small file on your site called robots.txt decides which of these bots are allowed in.
Here is the trap: many security plugins, privacy tools, and hosts block AI bots by default, and they rarely tell you. If PerplexityBot is blocked, Perplexity can never cite your page, with no error to warn you.
| AI bot | Vendor | Role | If you block it |
|---|---|---|---|
| OAI-SearchBot | OpenAI | Retrieval | ChatGPT Search can no longer cite you |
| ChatGPT-User | OpenAI | Retrieval | ChatGPT browsing can no longer read the page |
| PerplexityBot | Perplexity | Retrieval | Perplexity cannot crawl or cite you |
| Bingbot | Microsoft | Retrieval | Bing and Copilot lose access |
| GPTBot | OpenAI | Training | Opts your content out of model training |
| Google-Extended | Training | Opts you out of Gemini training | |
| CCBot | Common Crawl | Training | Opts you out of the Common Crawl dataset |
Notice the two roles. Retrieval bots fetch your page live to answer a question right now, so blocking one removes you from that tool's answers. Training bots collect data to train future models, and blocking those is a valid opt-out that does not hurt your live visibility.
# Let AI answer engines read and cite you
User-agent: OAI-SearchBot
Allow: /
User-agent: ChatGPT-User
Allow: /
User-agent: PerplexityBot
Allow: /
# Opt out of model training (optional)
User-agent: GPTBot
Disallow: /Want a quick look without a tool? Open your robots.txt in a browser, or pull it from the command line:
# Read your live robots.txt
curl -s https://example.com/robots.txt
# Check whether GPTBot is mentioned at all
curl -s https://example.com/robots.txt | grep -i gptbotFor the exact rules, see the official robots.txt guide, plus each vendor's bot docs from OpenAI, Perplexity, Anthropic, and Common Crawl.
llms.txt: a simple map for AI models
llms.txt is a small text file you place at the root of your site. Think of it as robots.txt, but made for large language models instead of search crawlers.
Where robots.txt says which bots are allowed, llms.txt points AI tools to your most important, cleanest content in plain Markdown. It is a way to say, in one short file, "these are the pages worth reading."
# Example Co
> Running-shoe reviews and buying guides.
## Guides
- [Best running shoes 2025](https://example.com/best-running-shoes): tested over 600 miles
- [How to pick trail shoes](https://example.com/trail-shoes): fit, grip and drop explainedIt is a newer idea, and not every engine has confirmed it reads the file yet, so the checker treats it as a low-effort bonus rather than a must-have. You can read the full proposal at llmstxt.org. Because it costs almost nothing to add, we flag it when it is missing so you can decide.
A longer companion file, llms-full.txt, can hold the full text of those pages in one place. It is optional too, but handy if you want AI tools to have your content in a single clean document.
Can AI read the page without JavaScript?
Letting a bot in is only half the job. It still has to read your words.
A web browser runs JavaScript to build the page you see. But many AI crawlers grab the raw HTML and skip that step. If your main text only appears after JavaScript loads, those bots may see an almost empty page.
Content that often hides behind JavaScript:
- Text tucked inside tabs or accordions that only open on click
- Reviews, comments, and ratings pulled in after the page loads
- Posts hidden behind a "load more" button
- Whole pages built by a single-page app framework
The safe rule is simple: keep your main text in the HTML, not locked behind scripts. Google explains the same problem for its own crawler in its JavaScript SEO basics. Our free JavaScript SEO checker goes deeper on the exact render gap if a check flags it.
Structure and metadata: helping AI trust you
Once a bot can read the page, clear structure helps it understand and quote you.
Structured data is hidden code, usually JSON-LD, that labels what your page is about in a language machines understand. It says, in plain terms, "this is an article, written by this person, on this date." Adding the right type, like Article or FAQPage, is exactly the kind of signal an AI uses to decide if your page is trustworthy enough to cite. You can explore the types at schema.org and read Google's rules in its structured data guide.
On top of that, the checker looks at the basics any engine reads first: a clear title, a meta description, and exactly one H1. It also looks for question-style headings, because AI tools fan a topic out into many small questions, and headings that match those questions are easier to pull a direct answer from.
What a URL fetch cannot measure
An honest readiness check tells you what it cannot see, not just what it can.
A single fetch reads your public files: robots.txt, the HTML, your sitemap, your schema. It cannot read your server logs, so it cannot prove that GPTBot or PerplexityBot actually visited. And it cannot sample thousands of AI answers, so it cannot tell you whether ChatGPT already cites you today.
Those need server-side data and answer sampling, which a free tool flags instead of faking a number. What the free check does well is catch the quiet blockers, a disallowed bot, a JavaScript wall, a missing schema, that stop you from ever being read in the first place.
Fix it in order, then re-check
The fastest path to AI visibility is boring, and that is good news.
Start at the top: unblock the retrieval bots in robots.txt. Then make sure your main content is in the raw HTML. Then add the schema that fits the page, and a short llms.txt if you want the bonus signal. Each fix is small, and the order matters, because access comes before everything else.
Run the check above on your own URL to see where you stand, then re-check after each change to confirm the fix landed.
Frequently asked questions
- What is an AI readiness checker?
- An AI readiness checker tests whether AI tools like ChatGPT, Perplexity, and Google AI Overviews can find, crawl, read, and cite your page. It reads your robots.txt, lists each AI bot, and shows which are allowed or blocked. It also checks the signals those engines rely on, like an llms.txt file, structured data, a clean title, and whether the page can be indexed at all.
- What is llms.txt and do I need it?
- llms.txt is a small text file you place at the root of your site, like robots.txt but aimed at large language models. It points AI tools to your most important, clean content in plain Markdown. It is optional and not yet read by every major engine, but it is a low-effort signal that can help AI find and quote your best pages. The checker flags whether you have one so you can decide.
- How do I let ChatGPT and Perplexity crawl my site?
- Edit your robots.txt to allow their retrieval bots. ChatGPT Search uses OAI-SearchBot and ChatGPT-User, and Perplexity uses PerplexityBot. Make sure none of these are disallowed. Many security plugins and hosts block AI crawlers by default, so check your file even if you never wrote a rule yourself.
- What is the difference between retrieval bots and training bots?
- Retrieval bots fetch your page live to answer a question right now, so blocking one removes you from that tool's AI answers. Training bots collect data to train future models, and blocking those is a valid opt-out that does not hurt your live AI visibility. The checker splits the two so you only block what you mean to.
- Should I block GPTBot?
- It depends on your goal. GPTBot is mostly a training crawler, so blocking it opts your content out of model training without removing you from ChatGPT Search, which uses OAI-SearchBot. If you want maximum AI visibility, allow the retrieval bots. If you only want to opt out of training, block GPTBot and keep the retrieval bots open.
- Why is my page not showing up in AI answers?
- The usual reasons are a blocked retrieval bot, content that only loads after JavaScript runs, missing structured data, or a page that is not indexable. The checker tests each of these and tells you which one is holding you back, with the fix for each, so you can act in order of impact.
- How is AI readiness different from normal SEO?
- Normal SEO mostly optimizes for Google's ten blue links: titles, links, speed, and indexing. AI readiness adds the AI layer on top: are AI crawlers allowed in your robots.txt, is your content readable without JavaScript, do you have an llms.txt and clear schema, and can an engine quote a direct answer from your page. The goal is to be cited in AI answers, not only ranked.
- Is the AI readiness checker free?
- Yes. You can run a set number of checks per day with no login. A free account unlocks the full audit, including the JavaScript render delta that shows what AI crawlers actually see, an AI-written assessment with ranked fixes, downloadable reports, an API, and automatic 14-day re-checks across all your sites.