Does Google's AI Overview cite you?

An AI Overview answers the searcher directly and names a few sources. If it appears for a search your buyers make, the ranking underneath it matters less than it used to, because a large share of those searches now end without a click on anything, and being one of the cited sources is the thing worth having. What this check gives you is the citation list for one search, right now. What it cannot give you is a trend, a score, or a fact that will still be true this afternoon. AI Overviews change between identical searches, and any tool that reports a single reading as a standing measurement is selling you a number it cannot support. The honest use of this page is to run it two or three times on the same search and watch what moves. That takes a minute and teaches you more about AI visibility than any dashboard will.

The AI-written summary Google places above the ordinary results on some searches. It answers the question directly and cites a handful of sources, which means a search that shows one can be answered without anybody clicking a result at all. Whether you are among those cited is a different question from where you rank.

Because AI Overviews are not stable. They vary by location, by device, by signed-in history, and they change between identical searches minutes apart. That is the single most important thing to understand about this surface, and it is why a tool that reported one check as a standing fact would be lying to you. Run it twice on purpose.

No. Google chooses to summarise some searches and not others, and that choice is about the query rather than about any particular site. No Overview means there was nothing to be cited in, for anyone, on that search at that moment.

Every other free tool here is plain HTTP against a page, which costs us nothing. Google publishes no API for AI Overviews, so this one reads a paid search API at roughly a cent and a half per call. Five a day is enough to check your own search, a competitor's, and to re-run one and watch it change, which is the point of the page.

No, and it is worth being precise. This checks one surface, on one query, once. ChatGPT, Claude, Perplexity and Gemini each decide separately, and each of them is non-deterministic too. Measuring visibility properly means asking a fixed set of buying questions repeatedly across engines and reporting the range the answer falls in.

Not with this page, and that is deliberate. Every reading here is one observation of one search at one moment, and AI Overviews change between identical searches, so a chart built from single readings would show movement that is mostly the engine talking to itself. Monitoring means running the same searches on a schedule, enough times each to know the spread, and reporting the band the number is accurate to alongside the number. That is what Track does, and it watches ChatGPT, Claude, Gemini and Perplexity as well as Google, because a buyer who asks one of them is not asking the others.