Google recently opened up its Notebooks feature inside Gemini to all free users on the web. It was previously available only to paid subscribers, and the rollout to everyone marks a real change in how accessible this kind of persistent AI workspace has become.
You might be aware of the “it’s free now” angle. That is the least interesting part of the story.
What Notebooks Actually Are?
Think of a Notebook as a project folder that Gemini remembers. Instead of starting every conversation from zero, you group your sources, files, and previous conversations under one topic. Gemini then reads everything inside that notebook when you ask it a question, so your responses get progressively more relevant as you add more material over time.
You find Notebooks in the left panel of Gemini on the web, sitting just above Gems and Chats. Any conversation you have in Gemini can be pulled into a notebook using the three-dot menu on that chat. Once inside, you can also set custom instructions to tell Gemini how you want it to respond inside that specific notebook, covering things like tone, format, and level of detail.
If you want Gemini to answer only from what is in the notebook rather than drawing on its general knowledge alongside your sources, there is a notebook memory toggle you can turn off.
The Source Limits You Should Know Before You Start
Free users can add up to 50 sources per notebook. AI Plus subscribers get 100 per notebook. Pro users get 300. Ultra subscribers get up to 600.
Fifty sources are more than enough for most personal projects. Where it starts to matter is if you are trying to upload an entire semester of lecture notes, a full research database, or years of company documents. In those cases, the free limit will feel tight quickly.
The NotebookLM Connection Is the Real Story
Gemini Notebooks and NotebookLM now share the same underlying workspace.
What that means in practice: you add sources in Gemini, then open NotebookLM, and those same sources are already there. You can then use NotebookLM’s specific output features, including Video Overviews, Infographics, Audio Overviews, and Study Guides on material you gathered while working in Gemini.
These two tools were previously separate. Now they are effectively two front-ends pointing at the same project. This is practically useful because the two tools are strong at different things. Gemini handles open-ended drafting, web search, and back-and-forth conversation well. NotebookLM is built for deep synthesis of uploaded documents and structured outputs like podcast-style audio summaries from your sources. Using both on the same notebook without manually transferring anything between them changes the workflow in a meaningful way.
What Notebooks Do Not Do That You Might Assume They Do
This is the part worth understanding before you invest time building notebooks.
Notebooks do not carry memory across projects. There is no continuity between your separate notebooks. If you want Gemini to know that you prefer bullet-point summaries, or that you are writing at a certain technical level, or that you work in a specific field, you need to set those custom instructions inside every notebook individually. Nothing carries over automatically from one notebook to another.
There is also a deletion detail worth knowing. If you delete a notebook, the notebook itself is deleted along with all its sources and NotebookLM outputs. However, your Gemini chat history from conversations inside that notebook is not deleted. Those chats get moved back into your main Gemini chat list. For most personal users, this will not matter. For anyone using notebooks for work-related projects with data retention considerations, it is a detail worth being aware of before deleting anything.
Gemini can read content from Google Docs and Drive when you add them as sources. But it cannot write back to those documents automatically from within a notebook. The connection is read-only in the current version.
Read More: Google Just Made This Common Website Trick a Spam Violation
A More Useful Way to Think About the 50-Source Limit
Most people will use this incorrectly initially. They will treat the notebook as a filing cabinet and try to add everything they have on a topic.
A more practical approach is to be selective. Each source you add shapes how Gemini responds inside that notebook. Adding ten well-chosen, directly relevant sources produces better outputs than adding fifty loosely related ones. Gemini works with everything in the notebook simultaneously, and the quality of its responses depends on the quality and relevance of what you have stored there.
One structure that works well in practice: keep one notebook per active project with only the sources that are directly relevant to what you are currently trying to produce. When the project shifts direction, update or replace sources rather than accumulating indefinitely.
The better your notebook is organised, the better Gemini performs inside it. This is not automatic. It is something you build deliberately over time.
What This Feature Is Really About
Google’s stated long-term goal is for Notebooks to become personal knowledge bases that work across Google products. The current version is a useful workspace. What it is designed to eventually become is something closer to a persistent layer of context that follows you across your Google account rather than resetting with each new conversation.
The shift from treating AI as something you query to treating it as something that knows your ongoing work is the direction this feature is pointing. Whether you experience that shift depends entirely on how deliberately you build your notebooks.
Where to Find It
Notebooks is live now at gemini.google.com on the web. Look for the Notebooks section in the left panel of the interface. For Enterprise users, the feature is off by default and needs to be enabled by an administrator in organisation settings.
The feature has not yet reached the mobile app or Mac desktop app. Google has confirmed broader availability is expected in the coming weeks.
Also Read: Google Just Made NEET Prep Free – But There’s a Catch