Can NotebookLM write papers? How it differs from Acadwrite
NotebookLM (now Gemini Notebook) and Acadwrite both work from sources; they split on where sources come from and what you get at the end. NotebookLM reads only what you add and returns audio overviews, flashcards and mind maps, good for understanding and revision; Acadwrite searches scholarly databases for papers you do not have yet and returns reviews and introductions with a reference list.
Product details checked against Google's official help center and plans page on 22 October 2026. Numbers in brackets link to the sources at the end of the page.
Choose by the job in front of you
| What you need to do | Start with | Why |
|---|---|---|
| Turn a pile of material into podcasts, flashcards and mind maps for revision | NotebookLM | Generates audio and video overviews, reports, flashcards, quizzes, mind maps and slides.[3] |
| Ask questions only about your own material, with answers pointing back to it | NotebookLM | Answers carry inline citations to your sources; the free plan holds 50 sources per notebook, each up to 500,000 words with no page limit.[2][1][4] |
| You need papers you do not have yet | Acadwrite | NotebookLM is not a scholarly database and discovers sources from the web and cloud documents; Acadwrite searches journal articles, arXiv and PubMed in Chinese and English from 4–8 angles.[5] |
| Write a review or introduction with a reference list, exported to Word | Acadwrite | NotebookLM says nothing about citation formats or export; Acadwrite confirms the paper list and outline, cites at least 30 papers and builds the reference list with the text.[3] |
| Understand a difficult paper and how far its evidence goes | Acadwrite | Paper X-ray separates load-bearing design from decoration and works key equations with real numbers; critical appraisal grades evidence with GRADE. |
| Simulate peer review or polish to thesis conventions | Acadwrite | Three simulated reviewers anchor comments to the text; thesis polishing diagnoses or rewrites against more than 60 rules. |
Side by side
Grey text means the vendor's pages do not say. That does not mean the feature is missing.
| NotebookLM and Acadwrite compared | NotebookLM | Acadwrite |
|---|---|---|
| Where sources come from | Sources you upload or add: PDFs, documents, web pages, YouTube, audio; source discovery and Deep Research find web pages[4][5] | Your uploaded PDFs plus Chinese and English papers searched from journals, arXiv and PubMed |
| How much it holds | 50 sources per notebook free, 300 on Pro; up to 500,000 words or 200 MB per source, no page limit[1][4] | A 1-million-token context |
| How answers cite | Inline citations to your own sources[2] | Every citation links to the paper, with the reference list built alongside |
| What you get | Audio and video overviews, reports, flashcards, quizzes, mind maps, slides[3] | Paper lists, outlines, reviews and introductions, edited in the workspace and exported to Word |
| Citation format and export | Not stated | Reference list generated with the text, exportable to Word |
| Usage and pricing | Free; personal accounts get compute-based quotas refreshed every 5 hours; Google AI Plus $4.99 a month, Pro $19.99[6][7] | Pay per use at ¥2 per credit; a review in a workspace uses about 8–9 credits |
| Academic features | Not stated | 23 academic skills, including close reading, simulated review, thesis polishing and reference checking |
Skills for turning understanding into a paper
NotebookLM helps you understand the material you have; judging the strength of evidence and writing to journal and thesis requirements each have skills written for academic work. They come from open-source projects on GitHub; enable them in the skill market and use them in workspace chats.
- 23
academic skills from 9 open-source projects, covering writing, polishing, review, verification and topic design
- 50K
stars for nature-skills, source of 10 skills including close reading, simulated review and proposal writing
- 54K
stars for K-Dense scientific-agent-skills, source of 5 skills including critical appraisal and scientific writing
- 20K
stars for Humanizer-zh, source of the Chinese de-templating skill
Star counts as shown on GitHub on 22 October 2026.
Close reading and appraisal
Paper X-ray
Reconstructs the starting point and the authors' bet, separates load-bearing design from decoration, works key equations with real numbers and checks how far the evidence goes.
nature-skillsPaper reading card
Breaks a paper into 16 fixed sections: method modules, the evidence chain from experiments to claims, author-stated limits kept apart from critique, and testable ideas.
nature-skillsCritical appraisal of claims
Examines design, bias and confounding and statistical misuse, grading evidence with GRADE and Cochrane risk-of-bias frameworks.
K-Dense scientific-agent-skills
Writing it up
Evidence-based scientific writing
Never invents data or citations, applies CONSORT, PRISMA, STROBE and other reporting guidelines by study design, and structures each IMRaD section.
K-Dense scientific-agent-skillsResearch proposals
Drafts proposals around an evidence table, argument map and section contracts, self-scores on 8 dimensions and softens claims the evidence cannot support.
nature-skillsThesis polishing
For Chinese science and engineering theses: diagnoses or rewrites against more than 60 rules on wording, sentences, punctuation, structure and equations.
graduate-thesis-polish-and-write-skill
Before you submit
Nature simulated reviewused in paper review
Three independent reviewers assess originality, significance and technical soundness, with comments anchored to the text.
nature-skillsReference checkingused in reference checks
Verifies authors, title, year, volume, pages and DOI against authoritative bibliographic metadata, and never substitutes a similar paper.
nature-skillsChinese de-templating
Removes filler and stock patterns against 31 checkpoints while keeping facts, hedges and the author's voice; no promise about AI detectors.
Humanizer-zh
FAQ
NotebookLM quotas and plans change often. Check Google's official help center.
Can NotebookLM write papers?
It helps you understand material and generates reports and slides, but Google does not say it writes whole papers or mention citation formats or export. It answers only from the sources you add, so literature beyond them has to be found separately.
Is NotebookLM good for literature reviews?
Good for reading: put dozens of papers in one notebook and ask how they agree or differ, with answers citing your sources. A review still needs two more steps: finding the papers you do not have, and writing them up with a reference list.
How much can the free NotebookLM plan hold?
Personal free accounts get up to 50 sources per notebook and 100 notebooks; each source can be up to 500,000 words or 200 MB with no page limit. Usage is compute-based, refreshing every 5 hours up to a weekly limit.
What is the relationship between NotebookLM and Gemini Notebook?
They are the same product. Google's pages now use the name Gemini Notebook, and the old NotebookLM help center addresses still work.
Can I use Acadwrite and NotebookLM together?
Yes. Use NotebookLM to understand your material and make flashcards for revision; use Acadwrite when you need more papers, a written review or introduction, or a pre-submission review. Acadwrite charges by use at ¥2 per credit.
Sources
All pages accessed on 22 October 2026.
- [1]Google：Gemini Notebook help: personal plans and source limits
- [2]Google：Gemini Notebook help: inline citations and languages
- [3]Google：Gemini Notebook help: outputs and quotas
- [4]Google：Gemini Notebook help: supported source types and sizes
- [5]Google：Gemini Notebook plans
- [6]Google：Gemini Notebook help: compute-based usage
- [7]Google：Google AI plans