Gemini Notebook (formerly NotebookLM, renamed July 2026) is a standalone app at notebook.google.com that only answers from documents you explicitly upload, with a citation on every claim, and produces Studio outputs like Audio Overviews, Reports, and Mind Maps that nothing else on this page can make.
Gemini is Google's general-purpose assistant at gemini.google.com: broad knowledge, live web search, open-ended writing and coding, no source restriction.
Notebooks in Gemini is a third, separate thing: a feature launched inside the Gemini app itself in April 2026 that lets you organize chats and files into a project-style workspace. It syncs with Gemini Notebook two ways, but it's not the same product, and it can't produce Studio outputs either.
Same underlying Gemini models power all three. What differs is what each one is allowed to answer from, and what it can build for you once it has an answer.
A naming note: until July 2026 this comparison was called "NotebookLM vs Gemini." Google renamed NotebookLM to Gemini Notebook, then, separately and three months earlier, had already shipped a feature called Notebooks in Gemini inside the Gemini app. Those two facts together are why this is now a three-way disambiguation, not a two-way one, and why even some 2026 coverage of the rename gets the third product wrong or ignores it.
The confusion is understandable, and it's Google's doing
Gemini Notebook actually runs on Google's Gemini model under the hood, which is exactly why the names get confused, now more than ever, since Gemini Notebook used to be called NotebookLM. But "built on Gemini" and "is Gemini" are different things. Gemini the assistant answers like ChatGPT does: broadly, from training data and the open web. Gemini Notebook restricts itself to a fixed set of documents you explicitly choose, and treats anything outside them as out of scope. Notebooks in Gemini, the third piece, is conversation-centric rather than document-centric: it's a way to keep a set of Gemini chats and attached files organized as a project, without the citation-locked, sources-only behavior that defines Gemini Notebook.
We checked the old domain ourselves rather than take a secondary source's word for it: as of 30 August 2026, notebooklm.google.com returns an HTTP 301 redirect straight to notebook.google.com, with cache-control headers explicitly set to no-store, meaning Google wants every request to hit the new domain fresh rather than serve a cached old response. The shorter notebooklm.google alias does the same, redirecting to notebook.google. If you have the old URL bookmarked, it still works, it just won't stay on the old address.
Where each one wins
- Gemini Notebook wins for research over your own material: papers, contracts, meeting transcripts, course readings, anything where the answer has to trace back to a specific document with a citation you can click and verify. It's also the only one of the three that can produce Studio outputs, Audio Overviews, Video Overviews, Reports, Flashcards, Quizzes, and Mind Maps, none of which exist inside Gemini or Notebooks in Gemini.
- Gemini wins for open-ended writing, coding help, brainstorming, and questions about current events or general knowledge that no single document you own would answer. It can also pull in live web search and other connected tools mid-conversation, something Gemini Notebook deliberately doesn't do.
- Notebooks in Gemini wins when you want Gemini's open-ended, web-connected behavior but organized as a persistent project, say, a running conversation plus a folder of reference files, rather than starting fresh each chat. It's the middle ground: less restrictive than Gemini Notebook, more organized than a plain Gemini chat history.
A test that settles most cases
Could a person answer correctly from only the documents you would hand them? If yes, that's a Gemini Notebook job. If the honest answer needs knowledge or search beyond any specific document, that's a Gemini or Notebooks-in-Gemini job, and the choice between those two comes down to whether you want the citation-locked structure of a dedicated notebook or the flexibility of an ordinary chat with files attached. Full breakdown of Gemini Notebook against a general chatbot, including a feature-by-feature table, is in Gemini Notebook vs ChatGPT; the same logic applies here since Gemini and ChatGPT occupy the same category relative to Gemini Notebook.
How to actually attach a Gemini Notebook notebook inside Gemini
This is the one piece of real integration between the three products, and it shipped with Notebooks in Gemini in April 2026, three months before the rename. Inside a Gemini chat, typing @Gemini Notebook invokes the connected-app mention pattern; if you haven't linked the two yet, Gemini prompts you to connect them. Once attached, notebooks created in either surface, Gemini Notebook itself or Notebooks in Gemini, sync to both sidebars: sources, custom instructions, and renames carry over. The key limitation to know before you rely on this: a notebook attached inside a Gemini conversation can still pull in web search and other tools alongside your sources, so answers aren't citation-locked the way they are inside Gemini Notebook proper, and Studio outputs (Audio Overviews, Mind Maps, and the rest) remain exclusive to Gemini Notebook itself. You can reference your research inside Gemini, but you can't generate a Mind Map there.
Pricing: one subscription covers both
There's no separate Gemini Notebook subscription. A single Google AI plan raises your limits on Gemini Notebook, Gemini, and Notebooks in Gemini simultaneously.
Plan Price Storage Gemini Notebook limits (notebooks/sources/chats-day)
--------------------------------------------------------------------------------
Free $0 15GB 100 / 50 / 50
Google AI $4.99/mo 400GB 200 / 100 / 200
Plus
Google AI $19.99/mo 5TB 500 / 300 / 500
Pro
Google AI $99.99/mo 20TB 500 / 500 / 2,500
Ultra 20TB
Google AI $199.99/mo 30TB 500 / 600 / 5,000
Ultra 30TB
Source: support.google.com/gemininotebook/answer/16213268 and
one.google.com/about/google-ai-plans/, verified 30 Aug 2026. Full
tier-by-tier breakdown, including Audio/Video Overview and Deep
Research limits, in our dedicated limits guide.Full tier-by-tier numbers, straight from Google's own support page, including the Audio/Video Overview and Deep Research allowances this table leaves out, are in our complete limits guide. Note that these flat per-tier numbers are set to be replaced by a compute-based system starting 2 September 2026, covered in our dedicated article on that change.
The privacy difference nobody else covers
One real, practical asymmetry between the two chat surfaces: conversations inside Gemini Notebook are protected from being used to train Google's models regardless of your account's general activity settings, since the product is built around handling documents you may not want feeding a model. The same underlying question asked instead through the regular Gemini app or through Notebooks in Gemini is subject to your ordinary Gemini Apps Activity setting, meaning it can be used for training unless you've turned that setting off. If you're working with anything sensitive, client material, unpublished research, internal documents, doing it inside Gemini Notebook rather than pasting the same content into a Gemini chat is the safer default, not just a workflow preference.
When NOT to use each one
- Don't use Gemini Notebook for anything that needs current information beyond your uploaded sources, it will refuse or hedge rather than reach for the open web, by design. It's also the wrong tool if you need a quick, disposable answer; the overhead of adding sources first only pays off for material you'll return to.
- Don't use plain Gemini when you need every claim traceable to a specific document you can hand to someone else as proof; Gemini will blend web knowledge and uploaded files without the same citation discipline.
- Don't use Notebooks in Gemini expecting Gemini Notebook's Studio outputs, Audio Overviews, Mind Maps, Flashcards, none of that exists on this surface. It's an organizational layer over ordinary Gemini chat, not a lighter version of Gemini Notebook.
- A real failure mode worth knowing about from hands-on testing: asking Gemini (with a notebook attached) a document-grounded question can produce an answer that glances at your sources but drifts into the model's own general knowledge anyway, since attachment doesn't enforce the same lock Gemini Notebook does natively. If strict source-grounding is the entire point of the task, do it inside Gemini Notebook itself, not through the attachment.
Known rough edges
- No notebook-merge feature. If you end up with duplicate or fragmented notebooks across the two surfaces, the only fix is manually recreating or re-sharing content; there's no built-in merge.
- Shared notebooks don't carry into Gemini at all, by design. Google's own support documentation is explicit: any notebook you've shared with collaborators, even ones you own, will not be visible inside Gemini. And a notebook that contains Gemini chats specifically can't be shared at all. Both are hard rules, not sync lag.
One more adjacent mix-up: Deep Research
A separate but related confusion: Gemini has its own Deep Research mode, and Gemini Notebook has a different Deep Research mode, and they don't do the same thing. Gemini's version researches the open web. Gemini Notebook's version researches only within your notebook's existing sources. Our Gemini Notebook Deep Research guide covers the notebook-scoped version specifically; if you want the open-web version compared against ChatGPT, Claude, and Perplexity's equivalents, that's a different article entirely.
Using them together
They're not really competitors. A workflow that uses all three: verify and synthesize facts in Gemini Notebook, where citations keep every claim honest; pull that verified notebook into a Gemini conversation via @-mention when you need to combine it with live web search or ask something the sources alone can't answer; and use Notebooks in Gemini to keep that whole thread, plus any loose reference files that don't warrant a full Gemini Notebook notebook, organized as one project instead of scattered across your chat history.
People also ask
Is Gemini Notebook part of Gemini?
They're separate products that both run on Google's Gemini model and share one subscription. Gemini Notebook (formerly NotebookLM) is a standalone tool for document-grounded research; Gemini is the general-purpose assistant. They can be connected within the Gemini app via an @-mention.
What's the difference between Gemini Notebook and Notebooks in Gemini?
Gemini Notebook is the standalone, document-grounded research product (formerly NotebookLM) that only answers from sources you upload and produces Studio outputs like Audio Overviews and Mind Maps. Notebooks in Gemini, launched April 2026, is a feature inside the Gemini app for organizing chats and files into a project; it doesn't enforce source-only answers and can't produce Studio outputs.
Can Gemini do what Gemini Notebook does?
Not natively. Gemini can read files you attach to a conversation, including an attached Gemini Notebook notebook via @-mention, but it doesn't enforce the same citation-locked, sources-only behavior, and it can't produce Gemini Notebook's Studio outputs.
Does notebooklm.google.com still work?
Yes, it redirects. We checked directly on 30 August 2026: notebooklm.google.com returns an HTTP 301 redirect to notebook.google.com with no-store caching, so the old address still resolves but always forwards to the new one.
Are Gemini Notebook conversations more private than regular Gemini chats?
Yes, in one specific way. Gemini Notebook conversations are excluded from model training regardless of your account's general activity settings. The same question asked in the regular Gemini app or Notebooks in Gemini is subject to your ordinary Gemini Apps Activity setting and can be used for training unless you've turned that off.
Which one should I use for studying?
Gemini Notebook, for material grounded in your actual course readings and notes. See the student workflow guide for a full study setup.