What the checker tests
Give it your address and it reads your homepage and the files an engine looks for: robots.txt, the sitemap, llms.txt, the structured data, the page as it renders without JavaScript. It runs about a hundred checks in half a minute and scores them by section:
- Crawl access. Which AI crawlers your robots file lets in, and which it blocks by accident. OAI-SearchBot, PerplexityBot, ClaudeBot, Google-Extended and the rest, each named.
- Discovery. Whether Google and Bing have the site, whether a public crawler has reached it, whether the sitemap and llms.txt exist and agree with the pages.
- Description. Whether every page describes you the same way: titles, descriptions, headings, the Organization markup.
- Quotability. Whether a sentence can be lifted out of each page and quoted, and whether the answer sits under the heading that asks the question.
- Speed. How the page loads on a phone on an Indian mobile network, from Google’s own measurements where they exist.
- Local presence. The address, hours and coordinates in the markup, the Business Profile, and whether they agree with the site.
- Agent readiness. The files and markup an AI agent reads to act on the site rather than describe it.
Every check says what it found. On the paid plans it also says why it matters and gives the code to paste, written for your site and the platform it runs on.
What the score means, and what it does not
The score is a ceiling, not a prediction. A site an engine cannot read cannot be recommended, so a low score caps you. A high score means nothing on your side is in the way. It does not mean an engine will name you, because that depends on what other sites say about you, and no checker that reads only your site can know that.
Other checkers give a single AI visibility score from a model of the web. This one fetches your pages and reports what it found on each, so a failing item is a fact about your site and not an estimate. The audit is the same hundred checks on the whole site on the paid plans: every page crawled, templates, titles, broken links, duplicate pages.
The second number: do the engines recommend you?
That one comes from asking. The free plan writes ten questions a buyer in your category would type and puts them to Gemini as a buyer in India, from any state or district you pick. You see who it recommended, whether it read your site first, and how it described you. The paid plans ask all six engines, ChatGPT, Gemini, Claude, Perplexity and Google’s AI Overviews and AI Mode, and give the number a name: Share of Model, the share of the answers to your buyers’ questions that recommended you, per engine.
Both numbers together tell you which problem you have. Readable and not recommended is a reputation problem, fixed on other sites. Not readable is a site problem, fixed by you this week.
How to run it
- Create a free account. No card.
- Enter the site. The check runs on its own and the score appears in about half a minute.
- Read the failing items. Fix what fails, then run it again: the free plan has four checks a month.
- Run the free look at the engines, and read what Gemini says about you to a buyer in India.
Nothing of your site is stored and nothing on it is changed. The checker reads the same public pages any crawler can.
Questions
Is the AI visibility checker free?
Yes. The free plan runs the hundred checks on your own site four times a month, shows the score by section and the name of every failing check, and asks Gemini ten questions about your brand once a month. No card is needed.
What is a good AI visibility score?
One where nothing fails under crawl access and quotability, because those are the checks that stop an engine reading you at all. A score in the nineties means the site is out of the way. It does not mean you will be recommended; that is the second number, from asking the engines.
Does the checker need access to my site or my analytics?
No. It reads the public pages any crawler can reach, and nothing is stored or changed. Google Analytics and Search Console can be connected later, read-only, to see the visits that arrived from AI answers.
What is the difference between an AI visibility score and Share of Model?
The score says whether the engines can read you. Share of Model says whether they recommend you: the share of answers to your buyers’ questions that named you, per engine. A site can score well on the first and zero on the second, which is normal for a new brand.
More guides
- How to rank on Gemini and be named in its answers
- How to rank on Perplexity and be cited in its answers
- How to appear in Google AI Overviews in India
- How to rank in Google AI Mode, and be cited there
- How to get ChatGPT to recommend your business
- Answer engine optimization services in India
- How to rank on ChatGPT, and what ranking even means there