How to check and track brand mentions in Gemini

The short answer

• For the Gemini app: check in a Temporary Chat, which Google does not use to personalize Gemini, open Sources when an answer shows them, and track the same questions through the Gemini API with Grounding with Google Search.
• For AI Overviews and AI Mode: use the generative AI performance report in Search Console, which shows impressions by page, country, device and date.
• Know your Google-Extended setting: it decides whether Gemini may use your content for model training and for grounding answers in the Gemini app. It does not change your Google Search ranking.

Three Google surfaces, three ways to measure

Measure each surface on its own terms. Impressions in AI Overviews say nothing about how the Gemini app describes your brand, and a good Gemini answer says nothing about your AI Overviews coverage.

What Google says about Gemini's sources

Google does not publish how the Gemini app chooses which brands to mention. It does document three things that matter for site owners:

• Grounding. Google describes grounding as supplying content from the Google Search index to the model at the time of the prompt, to improve factuality and relevance. A page that is not in Google's index cannot be part of that.
• Google-Extended. This is a robots.txt product token, not a separate crawler: Google crawls with its existing user agents. It lets you decide whether content Google crawls may be used to train future Gemini models and for grounding in Gemini Apps and in Grounding with Google Search on Vertex AI. Google says it has no effect on inclusion or ranking in Google Search.
• Sources in answers. Gemini Apps sometimes show sources and related content within and below a response, opened from a Sources button. Google says not every response includes them.

Step 1: Check in the Gemini app

1. Start a Temporary Chat. Google says temporary chats are not used to personalize Gemini or to train its models, and are kept for up to 72 hours.
2. If you check from your normal account instead, note that Google's announcement describes the personal context setting "Your past chats with Gemini" as on by default; it lets Gemini learn from earlier conversations.
3. Ask each question from your prompt set exactly as written, including local questions with the city in the wording.
4. Open Sources when it appears and note which of your pages are linked. Record answers without sources as their own category.
5. Ask the same question in Google Search with AI Mode and record that separately, because it is a different surface.

Keep the date, prompt, account state, whether you were mentioned and where, the competitors named and the linked pages for every check.

Step 2: Track it with the Gemini API

For repeatable samples, add the google_search tool to a Gemini API request. Older models use google_search_retrieval instead. The response includes:

• google_search_call, with the search queries the model ran;
• url_citation annotations on the answer text, each with a url, a title and the span of text it supports;
• search_suggestions, an HTML snippet whose display rules are set out in the Gemini API terms, should you show results to users.

For Gemini 3 and newer models, Google bills grounding per search query the model runs, with the first 5,000 search requests a month free and $14 per 1,000 after that, as of the documentation updated on 2026-10-13. Several queries for one prompt count as several uses, which adds up quickly across a weekly prompt set.

API answers will not match the app exactly, because the app adds personal context, its own settings and other features. Keep the model and wording fixed, and treat the API as a consistent instrument rather than a copy of what each user sees.

Step 3: Read AI Overviews and AI Mode in Search Console

Search Console added generative AI performance reports on 2026-06-03 and, by Google's note, made them available to all sites worldwide on 2026-08-31. The Search report shows how often your URLs appear in AI Overviews and AI Mode, broken down by page, country, device and date, down to hourly granularity. The same impressions also remain in the standard Performance report.

The report counts impressions only. It tells you which pages appear, not which brand an answer recommends or how it describes you, so pair it with prompt sampling.

Step 4: Decide on Google-Extended deliberately

Disallowing the token looks like this:

User-agent: Google-Extended
Disallow: /

That asks Google not to use your content for training future Gemini models or for grounding in Gemini Apps and Vertex AI. It does not remove pages from Google Search, AI Overviews or AI Mode; those follow Search controls such as nosnippet and max-snippet. If you want the Gemini app to be able to draw on your pages, leave Google-Extended allowed. Our AI crawler list shows how it sits beside the other engines' agents.

When Gemini leaves you out

• Is the page indexed? Check it with the URL Inspection tool in Search Console.
• Is it eligible for a snippet? Google requires pages to be indexed and eligible to show a snippet to appear as supporting links in AI Overviews and AI Mode.
• Is Google-Extended allowed? A rule that disallows it to stop AI training also stops grounding in the Gemini app.
• Does the page answer the question? Compare it with the sources Gemini linked instead. Structured data should describe the visible facts; see schema markup for AI search.
• Any technical blockers? Run the free AI visibility checker on the page.

CiteVisor's paid plans sample Gemini through the API each week alongside ChatGPT, Claude and Perplexity. The overview of tracking brand mentions in AI search covers the shared method, with separate guides for ChatGPT and Perplexity.

Related guides

• Track brand mentions in AI search with a fixed prompt set and separate mention and citation counts.
• Compare ChatGPT mentions and Perplexity mentions on the same questions.
• Use how to get cited in AI search and how to get cited by ChatGPT to plan improvements across engines.
• Review schema markup for AI search and the AI crawler list, then run the free AI visibility checker.

Sources

• Google Search Central: Google's common crawlers (Google-Extended)
• Google AI for Developers: Grounding with Google Search
• Gemini Apps Help: View related sources from Gemini Apps
• Google: Temporary chats and new personalization controls in Gemini
• Google Search Central Blog: Search generative AI performance reports
• Google Search Central: AI features and your website

Reviewed on 2026-10-21. Gemini features and API billing change often; check the linked pages before relying on a detail.