Updated on
NotebookLM: reviews and analysis
The research notebook that answers only from your own documents, with the citation attached.
Affiliate link · no extra cost to you
NotebookLM plays a different game from a chatbot: it answers only with the documents you upload and cites the exact passage, so verifying takes one click. The free tier holds up for real work and paid limits ride along with Google's subscriptions, from $4.99 a month. It falls short if you need an API, automation or ideas that aren't in your sources.
Best for: Anyone researching with their own corpus who needs every answer with a verifiable source attached.
What the internet says
The praise is always the same: it answers only from the documents you upload and cites the exact passage, so people working with their own material trust it more than a generalist assistant. The complaints repeat too, the 50-source cap per notebook on the free plan, the lack of an API and audio voices that wear thin by the third listen.
What the web repeats in favour
- Answers come only from the notebook's sources and every claim carries a citation that opens the original passage, the most cited reason for preferring it over a chatbot
- The free tier holds up for real work: 100 notebooks, 50 sources per notebook, 50 questions and 3 audio overviews a day at no cost
- Audio Overviews work in 80+ languages across four formats (deep dive, brief, critique and debate), and they are what people most recommend for reviewing dense material
- It fits professional work with a closed corpus, from legal files to internal regulations, and the iOS app sits around 4.9 out of 5 across some 51,000 ratings
What the web repeats against
- The free plan caps at 50 sources per notebook, so anyone with a large corpus ends up splitting it across notebooks or moving up a tier
- There is no public API and no automation, everything goes through the web interface one click at a time, which rules it out of any scripted workflow
- It tends to flatten technical material and to over-weight the first document in the notebook when sources contradict each other
- The audio voices sound like generic podcast hosts and interactive mode is English-only, which shows when you work in another language
- The July 2026 rename to Gemini Notebook revived distrust about the longevity of Google products, its third name since launch
Sweep sources: Official pricing · Docs · Docs · App stores · Communities · Communities · Press · Press
Latest NotebookLM news
NotebookLM becomes Gemini Notebook and gains code execution
Google keeps it as a standalone product and adds a cloud computer that writes and runs code inside each notebook, first for Ultra and Workspace and later for Pro. The rename leaves pricing and plans untouched, and existing notebooks and shared links keep working. Source
Pros / Cons
Pros
- Answers only from your sources and cites the exact passage
- Roomy free tier: 100 notebooks with 50 sources each
- Audio and video overviews in more than 80 languages
Cons
- No public API and no automation
- The free plan caps at 50 sources per notebook
- No use for writing from scratch or reasoning beyond your documents
TLDR: NotebookLM is a research notebook that answers only from the documents you upload, citing the exact passage behind every sentence. That rules it out of the open-ended chat ChatGPT is built for, and makes it the better pick when the material is yours and the source has to stay in view. The free tier holds up for real work, and paid limits ride along with Google’s subscriptions from $4.99 a month. In July 2026 Google renamed it Gemini Notebook.
What NotebookLM is and how it works
NotebookLM is Google’s research notebook. You create a notebook, upload your sources (PDFs, Drive documents, web pages, YouTube videos, audio recordings) and ask questions from there. The difference from a generalist assistant like ChatGPT or Gemini is where the answers come from. A chatbot replies with what it learned from the entire world, and you are left wondering whether it read that or inferred it. NotebookLM replies with what sits inside your notebook and attaches a numbered citation that opens the original passage in the document.
That confinement is exactly why people pay for it. When you supply the material, the room for error shrinks to misreading a paragraph, and that paragraph is one click away for checking. It’s the difference between asking someone to summarise a topic and asking them to summarise this folder.
On top of that base, the Studio panel turns the notebook into working material: audio overviews where two voices discuss your documents, explainer videos, mind maps, flashcards and quizzes. Audio runs in more than eighty languages and comes in four formats, from the long conversation to a two-minute brief, a critique or a debate. Interactive mode, the one that lets you cut in on the hosts, is English-only.
Google renamed it Gemini Notebook in July 2026 and added an environment where it can write and run code to analyse the data in your sources. Notebooks and shared links stayed put, so the change touches the name and the logo more than the daily work. We call it NotebookLM here because that’s what everyone searches for.
What it’s like to use day to day
The learning curve is minutes long. Upload documents, wait for processing, ask in plain language. What changes the work is the citation column: every claim comes tagged, and clicking it opens the exact passage in the source. Verifying stops being a separate task, and on sensitive material that beats any showy feature.
The friction lives elsewhere. The free plan takes fifty sources per notebook, enough for one topic or one case, tight for a thesis, and anyone who hits that ceiling ends up splitting the project across notebooks. Generating an audio or video overview takes several minutes. And there is no public API, which means everything goes through the interface, one click at a time. If the plan was to process four hundred reports on a schedule, this isn’t the place.
It also helps to know how it fails. On technical material it tends to flatten nuance and hand back a simpler version than the original. When two sources contradict each other, it usually carries the dominant document’s line instead of flagging the conflict, so there you have to ask explicitly and have it compare what each source says. And the audio voices, convincing at first, turn monotonous by the third listen.
Pricing and plans
The model is freemium and the free tier isn’t a shop window. It gives you a hundred notebooks, fifty sources in each, a few dozen questions a day and three audio overviews daily, more than a normal user burns through in a week of studying. Plenty of people never need to pay.
When you do pay, you aren’t paying for the notebook. There is no standalone subscription: the higher limits arrive inside Google’s plans, from $4.99 a month, with the Pro tier raising the caps further and getting new features first. If you already pay Google for Gemini or for storage, this comes included and its marginal cost is zero. If all you want is the notebook, you are buying a whole bundle to use one piece.
Who it’s for (and who it isn’t)
It’s for people working with their own corpus who need to answer with the source in hand. PhD students with fifty papers, exam candidates with a fixed syllabus, lawyers cross-checking a case file against the applicable rules, analysts pulling conclusions from twenty reports, teachers building material from their own notes. In every one of those cases the verifiable citation is the product.
It isn’t for people who want to generate new ideas, write from scratch or reason about something that isn’t in their documents: Claude gives better long-form text and better conversation for that. It also isn’t for anyone who needs to automate, integrate or wire this into a scripted workflow, because without an API there is no way in. And if what you want is an assistant that also reads your mail, calendar and Drive, that’s Gemini’s job, not its document-bound sibling’s.
Alternatives to NotebookLM
The catalogue has no direct rival of the same species, because almost everything else is a generalist assistant that lets you attach files, which is a surface resemblance. The closest options are Gemini and Claude, both able to work with long documents, neither built around citations and a persistent notebook. The full comparison, including the cases where switching does make sense, is on the NotebookLM alternatives page.
For turning notes into quizzes and flashcards with more control there is Quizgecko, and for searching the open web rather than your own documents, Kagi. The wider picture is in our guide to the best AI productivity tools.
Frequently asked questions
Is NotebookLM free?
Yes, with a permanent free tier that’s good enough to work in: a hundred notebooks, fifty sources per notebook and several audio overviews a day. Higher limits come with Google’s subscriptions, from $4.99 a month.
Does it make things up?
Far less than a chatbot, because it can only answer from the sources in the notebook and tags each claim with its citation. The typical failure isn’t inventing a fact, it’s oversimplifying or sticking with one document’s version when two disagree. The citation is right there to check it.
Is it the same as Gemini now that it’s called Gemini Notebook?
No. Google consolidated its AI branding in July 2026 and changed the name and the logo, but it remains a standalone product and existing notebooks and shared links didn’t move. Gemini is the generalist assistant, with access to the web and your mail. NotebookLM works sealed inside the documents you upload.
Can I use it on my phone?
Yes, there are iOS and Android apps alongside the web version, with one of the highest App Store ratings in its category. On mobile you listen to audio overviews and ask one-off questions rather than build the notebook.
Best alternatives to NotebookLM
Guides that use NotebookLM
How to research your competitors with AI in one morning
Guide to AI competitor research: noise-free search with Kagi, source-verified extraction and…
How to run document due diligence with AI without losing control
Guide to mass document review with AI: the inventory that orders, the layered analysis with…
How to get ready for e-invoicing with AI and without panic
Guide to e-invoicing for SMBs: what really changes, how to choose a compliant platform and the…