Short answer
Google-GeminiNotebook is the user agent Google's Gemini Notebook uses when it fetches a web page that one of its users has added as a source. Gemini Notebook is the product formerly called NotebookLM, and Google lists Google-NotebookLM as the former user agent, supported until August 2026. Google classes it as a user-triggered fetcher, which generally ignores robots.txt.
Key takeaways
- It fetches a page because a Gemini Notebook user added it.
- Google-NotebookLM was the old name, supported until August 2026.
- User-triggered fetchers generally ignore robots.txt; verify by IP and DNS.
What is Google-GeminiNotebook?
Google-GeminiNotebook is the user agent Google's Gemini Notebook sends when it loads a web page for one of its users. Google's fetcher list says: "The Gemini Notebook fetcher requests individual URLs that Gemini Notebook users have provided as sources for their projects."
Gemini Notebook is Google's AI research tool. Its home page calls it "the AI research tool and thinking partner that can analyze your sources". A person adds sources to a notebook, including web pages, and Gemini Notebook works from them. Google's announcement gives examples such as students converting notes into audio and video summaries.
So when you see Google-GeminiNotebook in your logs, someone has pasted one of your URLs into a notebook. Gemini Notebook fetched that page so it could read it. It is a single visit, triggered by one person, for one page.
Google lists it with its user-triggered fetchers. Those are fetchers a person starts, as opposed to crawlers that run on their own schedule. Our glossary has a short explainer on the user-triggered fetcher.
Why did the user agent change from Google-NotebookLM?
Because the product changed its name. On 16 July 2026 Google announced: "We're renaming NotebookLM to Gemini Notebook." It called the result "the same standalone product, now doing more across the Google ecosystem".
Google's fetcher list now shows the new token and, under "Former agent (supported until August 2026)", the old one: Google-NotebookLM. In short:
| Name | Status on Google's fetcher list |
|---|---|
Google-GeminiNotebook |
Current user agent token for Gemini Notebook |
Google-NotebookLM |
Former agent, "supported until August 2026" |
If you wrote firewall, CDN or analytics rules that match Google-NotebookLM, they won't match the new name. Add Google-GeminiNotebook alongside the old token so your rules and reports keep working.
What does Google-GeminiNotebook look like in my logs?
It looks like a Chrome browser string with the Google-GeminiNotebook token added. Google publishes a mobile and a desktop version:
Mobile:
Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/138.0.0.0 Mobile Safari/537.36 (compatible; Google-GeminiNotebook; +https://developers.google.com/crawling/docs/crawlers-fetchers/google-gemininotebook)
Desktop:
Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36 (compatible; Google-GeminiNotebook; +https://developers.google.com/crawling/docs/crawlers-fetchers/google-gemininotebook)
Match the Google-GeminiNotebook token, not the whole string. The Chrome version numbers in Google's examples are the kind of detail that can change.
To find these visits, search your raw server or CDN logs for the token. Our guide to tracking AI agent traffic covers where to look.
Google puts a caution at the top of the list: "The user agent string can be spoofed." The token tells you what a request claims to be. The checks below tell you whether it is true.
Does Google-GeminiNotebook follow robots.txt?
Generally, no. Google says of all its user-triggered fetchers: "Because the fetch was requested by a user, these fetchers generally ignore robots.txt rules."
Google makes no separate statement for Gemini Notebook, so this general rule is the best guidance available. A Disallow line aimed at Google-GeminiNotebook is unlikely to stop it.
Your Googlebot rules don't control it either. Google puts Googlebot in its common crawlers category and Gemini Notebook in its user-triggered fetchers category. For how the different kinds of AI bots treat robots.txt, see robots.txt for AI agents.
How do I check a request is really from Google?
Use Google's published IP lists or a reverse DNS check. Google's verification page splits its user-triggered fetchers into two groups:
- Fetchers controlled by Google come from IPs in
user-triggered-fetchers-google.jsonand "resolve to a google.com hostname". - Fetchers on Google Cloud sites come from IPs in
user-triggered-fetchers.json, which resolve togae.googleusercontent.comhostnames. Google gives the example of a site on Google Cloud that fetches RSS feeds for its own users.
Google's fetcher page does not say, fetcher by fetcher, which list each one uses. It does say the reverse DNS mask for user-triggered fetchers matches ***-***-***-***.gae.googleusercontent.com or google-proxy-***-***-***-***.google.com.
For a one-off check, Google describes two steps:
- Run a reverse DNS lookup on the IP address from your logs and confirm the hostname ends in
googlebot.com,google.comorgoogleusercontent.com. - Run a forward DNS lookup on that hostname and confirm it returns the same IP address.
For regular checks, match the IP against Google's JSON lists, and fetch them on a schedule. Each file carries its own creation time and changes over time. If a request carries the Google-GeminiNotebook token but fails both checks, treat it as a request that claims to be Google-GeminiNotebook but fails verification. Our reference on IP ranges and reverse DNS shows how this works across AI companies.
Google-GeminiNotebook is not documented as signing its requests. Google mentions Web Bot Auth only for Google-Agent, which is a different fetcher for agents that take actions.
Should I allow or block Google-GeminiNotebook?
For most stores and business sites, allow it. Each visit means a real person chose your page as a source for their own research. They may be comparing suppliers, studying a product category or preparing a report that cites you.
Some reasons to be more careful:
- Pages behind a login. If content should only reach customers who are signed in, protect it with a login. Robots.txt won't stop a user-triggered fetcher.
- Paid or licensed content. If you sell access to your writing or data, use the same access controls you use for any visitor.
- A hard block. If you decide to block it, do so in your firewall or CDN by matching the
Google-GeminiNotebookandGoogle-NotebookLMtokens. Because a token can be spoofed, check the IP or DNS before trusting an allow rule.
How do I make my pages useful to Gemini Notebook users?
Write the page so it stands on its own. Someone added your URL as a source because they want to work from what it says. Pages that state key facts in plain text, such as prices, specifications, opening hours, delivery terms and return policies, give that person something solid to work from.
A few habits help any automated reader, not just this one:
- Put the answer near the top. Lead with the main fact, then add detail.
- Keep important text out of images. A price or policy shown only in a picture is harder for software to read reliably.
- Date your pages. If a policy or price changes, a visible "last updated" date tells the reader which version they are looking at.
- Keep one URL per topic. If the same information lives on several pages, people may add an out-of-date copy as their source.
Our llms.txt guide covers another way to give AI tools a clear map of your site.
For a wider decision guide across training crawlers, search bots and agents, see should you block AI agents?.
How Sill helps. Sill shows you which AI agents and bots visit your site and the pages they read. To see how ready your site is for AI agents, run the free scan.
Questions people ask
Is NotebookLM the same as Gemini Notebook?
Yes. Google announced on 16 July 2026 that it was renaming NotebookLM to Gemini Notebook, and called it the same standalone product.
Does Google-GeminiNotebook crawl my whole site?
Google says it requests individual URLs that Gemini Notebook users have provided as sources. That describes single-page fetches, not a crawl of your site.
Does Google-GeminiNotebook sign its requests with Web Bot Auth?
Google's fetcher list does not say so. The only user-triggered fetcher it mentions in connection with Web Bot Auth is Google-Agent, which Google says it is experimenting with.
Does blocking Google-GeminiNotebook affect my Google Search rankings?
Google's documentation does not say it does. Google documents Googlebot as a common crawler and lists Gemini Notebook separately as a user-triggered fetcher.
Sources
- Google: List of Google user-triggered fetchers (last updated 19 August 2026) · checked 17 Nov 2026
- Google: Verify requests from Google crawlers and fetchers · checked 17 Nov 2026
- Google: NotebookLM is now Gemini Notebook (16 July 2026) · checked 17 Nov 2026
- Google: Gemini Notebook · checked 17 Nov 2026
- Google: user-triggered-fetchers-google.json IP ranges · checked 17 Nov 2026