AI Readiness
How to make your website AI-ready: a guide to the AI Readiness Index
ChatGPT, Perplexity, Google AI Overviews and AI agents now read websites on their own. Here is what makes a site ready for them, how the AI Readiness Index scores it, and how to fix every gap - with a prompt your AI coding agent can run.
On this page

Key takeaways
- AI readiness is whether AI crawlers can reach a page, AI answer engines can understand and quote it, and AI agents can use it.
- The AI Readiness Index scores seven areas out of 100: crawler access, readable content, structured data, discovery files, GEO and AEO signals, agent readiness, and machine-readable agent interfaces.
- The most common blockers are a robots.txt or firewall that turns AI bots away, and pages that are empty without JavaScript.
- The free AI readiness checker gives you the score, the evidence and a prompt your AI coding agent can use to fix every issue.
- AI readiness complements a website audit and an SEO audit; it doesn't replace them.
People now ask ChatGPT, Claude, Perplexity and Google's AI Overviews what to buy and whom to hire, and AI agents are starting to browse and book on their behalf. None of them see your website the way a person does. They read the raw HTML, obey robots.txt, look for structured data and give up quickly. AI readiness is how well your site works for them - and you can measure it in about ten seconds with a free AI readiness checker.
Check your website's AI readiness now
What is website AI readiness?
A website is AI-ready when three things are true. AI crawlers can reach it: robots.txt, meta tags and the firewall let the bots in. AI systems can understand it: the content is in the HTML, clearly structured and described with schema.org data. AI agents can use it: links, buttons and forms are labelled, and the page is fast and secure. A website AI compatibility test checks all three at once.
This overlaps with SEO but isn't the same thing. Search engines render JavaScript and forgive a lot; most AI crawlers don't render JavaScript, and answer engines prefer content they can quote in one piece. That is why generative engine optimization (GEO) and answer engine optimization (AEO) have become their own disciplines.

The AI Readiness Index: seven areas, one score
The AI Readiness Index is the website AI readiness score our checker gives a page, from 0 to 100. It runs 48 checks grouped into seven areas, each area weighted by how much it affects AI visibility. Thirty-nine of the checks are scored; the other nine are reported with what we found but deliberately left out of the arithmetic, because they either only apply to some sites or can't be answered without running the page's JavaScript.
| Area | What it checks | Weight |
|---|---|---|
| AI crawler access | robots.txt rules for AI search bots (OAI-SearchBot, ChatGPT-User, Claude-SearchBot, PerplexityBot…) and training bots (GPTBot, ClaudeBot, Google-Extended…), noindex and nosnippet, and whether a firewall blocks AI user agents. | 23% |
| Content AI can read | Words in the server-sent HTML, title, meta description, one H1, heading structure, page language. | 18% |
| Structured data | Valid JSON-LD, Organization and WebSite entities, page-type schema (Article, Product, FAQPage…), Open Graph. | 14% |
| AI discovery files | llms.txt, XML sitemap, sitemap listed in robots.txt, canonical URL. | 13% |
| GEO and AEO signals | Question-style headings, an FAQ, lists and tables, content dates, author and publisher. | 9% |
| AI agent readiness | HTTPS, server response time, semantic landmarks, labelled form fields, named links and buttons, image alt text. | 13% |
| Machine-readable interfaces | The agent-facing documents an AI agent looks for: an API catalogue, OAuth discovery metadata, an MCP server card, agent skill files, a content-usage declaration, Markdown content negotiation and agent discovery records. | 10% |
85 or more means ready for AI, 65–84 almost ready, 40–64 partly ready and below 40 not ready yet. Because each finding shows its evidence - the blocked bot names, the word count, the status code - the score is easy to verify.

How to check your AI readiness in three steps
- Open the free AI readiness checker and enter your homepage or any important page. No account is needed.
- Read your AI Readiness Index and the score for each area. Start with failed checks: they carry the most weight.
- Copy the AI agent prompt from the result and paste it into Claude Code, Cursor, GitHub Copilot or another coding assistant that can edit your site. Then run the check again to confirm.
Fix 1: let AI search crawlers in
The most expensive mistake is a robots.txt that blocks AI search bots, often copied from a template. Our AI crawler accessibility checker separates two groups. AI search and answer bots fetch pages to answer a question right now - block them and you can't be cited. AI training bots collect text for future models; blocking them is a legitimate choice and only costs a few points.
User-agent: *
Disallow: /admin/
Disallow: /account/
Disallow: /checkout/
User-agent: OAI-SearchBot
User-agent: ChatGPT-User
User-agent: Claude-SearchBot
User-agent: Claude-User
User-agent: PerplexityBot
Allow: /
Disallow: /admin/
Sitemap: https://yourwebsite.com/sitemap.xmlFix 2: put your content in the HTML
Most AI crawlers don't run JavaScript. A single-page app that sends <div id="root"></div> and builds everything in the browser looks empty to them. Use server-side rendering or static generation - Next.js, Nuxt, Astro, or your CMS's normal page output - so headings, body text, prices and FAQs are in the page source. Then check the basics a SEO audit also covers: one clear H1, descriptive H2s, a title and a meta description.
Fix 3: describe your business with structured data
JSON-LD states facts in a form machines read without guessing. Every site should have an Organization (or LocalBusiness) and a WebSite node; add Article, Product, Service or FAQPage where they fit. Only use real data. See how to check structured data for testing it.
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@graph": [
{
"@type": "Organization",
"name": "Your Company",
"url": "https://yourwebsite.com/",
"logo": "https://yourwebsite.com/logo.png",
"sameAs": ["https://www.linkedin.com/company/your-company"]
},
{ "@type": "WebSite", "name": "Your Company", "url": "https://yourwebsite.com/" }
]
}
</script>Fix 4: publish llms.txt and a sitemap
llms.txt is a short Markdown file at your site root that gives language models a clean map of your most important pages. It's optional and still emerging, but cheap to add. Pair it with an XML sitemap listed in robots.txt - see robots.txt and sitemap.xml explained.
# Your Company
> Emergency plumbers in Leeds: leaks, boilers and blocked drains, 24/7, fixed prices.
## Key pages
- [Services](https://yourwebsite.com/services): what we fix and how it works
- [Pricing](https://yourwebsite.com/pricing): fixed prices and call-out fees
- [Contact](https://yourwebsite.com/contact): phone, hours and service areaFix 5: write for answer engines (GEO and AEO)
A GEO readiness checker or AEO score checker looks for content an AI can lift and quote. Phrase some headings as the questions customers ask, answer each in the first sentence below it, and use lists and tables for steps, features and comparisons. Add a short FAQ marked up as FAQPage, show when the page was last updated, and say who wrote it.
- Question heading: “How much does an emergency call-out cost?”
- Direct answer first: “A call-out costs £79, including the first 30 minutes.”
- Then the detail: a short list or table of what's included.
Fix 6: make your site usable by AI agents
AI agents click links, fill forms and compare options. An AI agent readiness checker - or agentic web readiness checker - looks at what they rely on: HTTPS, a fast response, header, nav, main and footer landmarks, a label on every form field, a name on every icon button, and alt text on images. Most of this is also basic accessibility, so the accessibility checker helps too.
Let your AI coding agent fix it
Every result from the AI readiness checker ends with a ready-made prompt. It names your site and score, lists each failed check with the evidence we found and the exact fix, and tells the agent to use your framework's conventions, never invent facts for schema or llms.txt, keep private areas blocked, and explain every change. Paste it into Claude Code, Cursor or GitHub Copilot, review the changes, deploy, and run the check again.
AI readiness vs a website audit vs an SEO audit
| Check | Question it answers | Tool |
|---|---|---|
| AI readiness check | Can AI search, answer engines and agents reach, understand and use my site? | AI readiness checker |
| SEO audit | Can search engines index my pages and show them well? | SEO checker |
| Website audit | What is holding my whole site back - SEO, speed, mobile, accessibility, security, links? | Website audit |
They share foundations, so fixing one usually helps the others. Start with the AI readiness check, then run the full website audit and an SEO audit - the AI readiness result links straight to both for the same page.
Run the free checks
AI Readiness Checker
Get your AI Readiness Index: AI crawler access, GEO and AEO signals, structured data, llms.txt and agent readiness, with a fix prompt for your AI agent.
Check AI readinessWebsite Audit
Check one page for SEO, speed, mobile, accessibility, security, broken links and server problems in a single report.
Audit my websiteSEO Checker
Check your page title, description, headings, image descriptions, canonical URL, robots.txt and sitemap.
Check SEOFrequently asked questions
What is an AI readiness checker?
A tool that tests whether a website is ready for AI search engines, answer engines and AI agents. Ours reads the page the way an AI crawler does and gives it an AI Readiness Index from 0 to 100, with a fix for every issue.
What is a good website AI readiness score?
85 or more is ready for AI, 65 to 84 almost ready, 40 to 64 partly ready and below 40 not ready yet. Fix failed checks first because they carry the most weight.
Does blocking GPTBot hurt my AI visibility?
GPTBot collects training data, so blocking it is a business choice and only lowers the score slightly. Blocking OAI-SearchBot or ChatGPT-User is different: those fetch pages to answer questions, so blocking them keeps you out of ChatGPT's answers.
Is llms.txt required?
No. It's an optional, emerging convention. It's quick to add and gives language models a clean summary of your key pages, so the checker counts it as part of AI discovery.
What is the difference between GEO and AEO?
Generative engine optimization (GEO) is about being used and cited in AI-generated answers such as ChatGPT or Google AI Overviews. Answer engine optimization (AEO) focuses on content that directly answers questions. The checker scores both together.
Is the AI readiness check free?
Yes. It needs no account and takes about ten seconds. It checks one page plus your site-wide files such as robots.txt, llms.txt and the sitemap.