Google-Extended

Google-Extended is a product token that site owners can use in robots.txt to control whether Google may use their content for Gemini. According to Google's crawler documentation, it covers two uses: training future generations of Gemini models that power Gemini Apps and the Vertex AI API for Gemini, and grounding, which provides content from the Google Search index to a model at prompt time, in Gemini Apps and in Grounding with Google Search on Vertex AI.

What it does not do

Google-Extended has no separate user agent string. Google crawls with its existing user agents, and the token works only as a control in robots.txt. Google states that Google-Extended "does not impact a site's inclusion in Google Search nor is it used as a ranking signal in Google Search." AI Overviews and AI Mode are part of Google Search, so blocking Google-Extended does not remove you from them. To limit what Google shows from your pages there, Google points to snippet controls such as nosnippet, data-nosnippet and max-snippet, and to noindex.

Why it matters for AI visibility

The token affects Gemini, one of the engines your buyers ask. Blocking it keeps your content out of grounding in Gemini Apps, so Gemini's answers there cannot draw on your pages through that route. One token covers both training and grounding, so decide on the two together.

How to set it

A robots.txt group for the token looks like any other:

User-agent: Google-Extended
Disallow: /

Leave the group out, or allow the paths you want used, if you want Gemini Apps to ground answers in your pages. Then track Gemini with prompt tracking to see whether its answers name and cite you. The equivalent bots from OpenAI, Anthropic and Perplexity are covered under AI crawler.