Free tool · AI & GEO
AI search readiness checker
How easy is your page for search engines and AI systems to fetch, understand, attribute and quote? Enter an address for a 0–100 readiness score across six areas, with the issues to fix first.
AI Search Readiness Checker
What it checks
Six areas that decide whether a page can be used
AI search products such as ChatGPT search, Perplexity, Claude and Google’s AI Overviews find pages through crawlers and search indexes, then extract passages to answer a question. The checker looks at the signals that make each step work, using the same public information those systems receive.
Crawlability
robots.txt rules for Googlebot, Bingbot, OAI-SearchBot, PerplexityBot, Claude-SearchBot and Applebot; noindex and nosnippet; robots.txt server errors; sitemap presence and whether the page is listed. Training crawlers such as GPTBot and Google-Extended are reported separately, because blocking them is a policy choice.
Technical
HTTP status, HTTPS, redirects, declared language, canonical tag, hreflang and mobile viewport, so the page is reachable and identified consistently.
Content
Title, description, heading structure, amount of text in the HTML, internal links, references to outside sources, and publication and update dates on articles.
Entities
Whether the organization behind the site is described in structured data with a name, URL and logo, linked to official profiles with sameAs, and whether contact details, an About page and article authors are present.
Structured data
Valid JSON-LD, the key properties of each type, a type that describes the page itself (Article, Product, Service, FAQPage and others), breadcrumbs and Open Graph metadata.
AI readability
Content available without JavaScript, question-and-answer sections, lists and tables, a short summary near the top, paragraph length, semantic HTML such as <main> and <article>, and plain definitions of key terms.
Why it matters
AI answers are built from passages
Generative search systems retrieve pages, split them into passages and quote or summarize the ones that answer a question directly. A page helps itself when a crawler is allowed to fetch it, the main content is present in the HTML, each section answers a clear question, and facts are stated plainly enough to quote without losing their meaning.
Attribution depends on entities. When structured data and visible content consistently say who publishes the page, who wrote it and how to contact them, systems can connect the page to a known organization or person instead of an anonymous URL. Search engines have long used the same signals.
Much of this is also good SEO and accessibility. Clear structure, server-rendered content and accurate metadata help search results, screen readers and translation tools as much as AI assistants.
An internal diagnostic, not a promise
The AI Search Readiness Score is HUB’s own weighting of these checks. No tool can see whether ChatGPT, Gemini, Perplexity or Google AI Overviews includes or cites a page, and no score guarantees that they will. Those systems decide based on the question, their index, other sources and signals that are not public. Use the score to find and fix technical and structural gaps.
How to fix problems
Make the page easy to fetch, then easy to quote
- 01Open access for the crawlers you want
Allow search and AI answer crawlers in robots.txt and check CDN or firewall bot settings, which can block crawlers that robots.txt allows. Decide separately whether to allow model-training crawlers.
- 02Put the content in the HTML
If the page renders its text with JavaScript, use server-side rendering or static generation for public pages.
- 03Answer questions directly
Open with a two- or three-sentence summary. Phrase subheadings as the questions your audience asks and follow each with a concise answer. Use lists and tables for steps, options and comparisons. The AI Content Optimizer suggests missing questions and headings.
- 04Say who is behind the page
Add Organization structured data with name, URL, logo and sameAs profiles; name authors on articles with a link to their profile; show dates. The Schema Generator creates the JSON-LD.
- 05Cite sources and keep facts current
Link to primary sources for figures and claims and update pages when facts change. Verifiable content is easier for people and systems to trust.
Questions
About AI search
readiness
Crawler rules are one part of readiness. For an online store, the AI Commerce Readiness Check also covers product data and trust signals.
What is GEO?
Generative engine optimization (GEO) is the practice of making content easy for AI-driven search and answer engines to find, understand and cite. Much of it overlaps with technical SEO and clear writing; the difference is more emphasis on passage-level answers, entities and crawler access for AI products.
Should I block GPTBot or other AI crawlers?
It depends on how you want your content used. Training crawlers (GPTBot, ClaudeBot, Google-Extended, CCBot) are separate from the crawlers that power search features (OAI-SearchBot, Claude-SearchBot, PerplexityBot). Many sites allow search crawlers and block training crawlers. The checker reports both without judging the training choice.
Do I need an llms.txt file?
No. llms.txt is an experimental proposal. Google has said it does not use it for Search, and no AI provider guarantees that it reads the file. The checker only notes whether one exists and does not score it.
Why is my score high but the page never appears in AI answers?
Readiness removes technical obstacles; it does not create demand or authority. Whether a page is chosen depends on the question, competing sources, the system’s own index and signals no outside tool can see.
Is anything stored?
The result is cached for 20 minutes, then deleted. HUB counts checks anonymously with the host name only. The optional AI action plan sends the findings, not the page text, to OpenAI.
Need help with what
the tools found?
HUB engineers implement technical SEO, performance, security and AI-readiness fixes on WordPress, Magento, custom PHP and modern JavaScript stacks.