How to Use NotebookLM in 2026: A Practical Guide to Google's Free Research Tool

NotebookLM in 2026: A Practical Guide for Research and API Documentation

NotebookLM turns a pile of documents into something you can question. Upload PDFs, docs, links, and videos, then chat with an AI that answers only from those sources—with clickable citations. In 2026, it can also generate podcast-style audio, narrated video explainers, and mind maps. For more on its research capabilities, see our Deep Research breakdown.

Try Apidog today

This guide covers setup, source limits, useful features, paid tiers, and a developer workflow for working with API documentation. It also reflects the product as of August 2026, based on Google’s official support documentation.

What is NotebookLM, and is it free?

NotebookLM is Google’s source-grounded AI research assistant. The standard tier is free and includes:

• 100 notebooks
• 50 sources per notebook
• 50 chat queries per day
• 3 Audio Overviews per day
• 3 Video Overviews per day
• 10 daily generations of reports, flashcards, quizzes, and mind maps

You need a Google account and a browser, or the mobile app for Android or iOS.

“Source-grounded” is the important distinction. A general chatbot may rely on training data and the open web, which can lead to outdated or fabricated answers. NotebookLM answers from your uploaded sources and cites each claim. If the sources do not answer a question, it says so instead of guessing.

Google has been folding NotebookLM into the Gemini brand during 2026. Official support pages now call it Gemini Notebook, but it remains the same product with the same limits. The app is still available at notebooklm.google.com, and most users still call it NotebookLM.

Set up your first notebook

You can create a useful notebook in about two minutes:

1. Open notebooklm.google.com and sign in. There is no waitlist or credit card requirement.
2. Select Create new notebook.
3. Give the notebook one focused purpose: a research topic, course, API, deal, or project. NotebookLM answers from everything in the notebook, so unrelated sources can reduce answer quality.
4. Add three or four initial sources.
5. Wait for NotebookLM to process and index them. Processing takes seconds to a few minutes, depending on file size.
6. Ask a specific question in the chat panel.
7. Check the citation numbers and click them to open the relevant passage in the source.
8. Pin useful answers to the Studio panel as notes.

The core workflow is simple: add sources, ask questions, verify citations, and save useful answers.

Add sources: supported formats and limits

Source quality determines answer quality, so keep these boundaries in mind.

Supported sources

NotebookLM accepts:

• PDFs and text files, including scanned documents
• Google Docs and Google Slides from Drive
• Public website URLs
• YouTube videos with captions
• Audio files, such as lectures and meeting recordings
• Markdown files
• Pasted text

For YouTube, NotebookLM reads the transcript rather than the video pixels.

Limits and common problems

The free plan supports 50 sources per notebook. Individual sources have long carried a limit of roughly 500,000 words or 200 MB per file. That is enough for a 900-page PDF, and usually more than enough for a typical project.

Watch for these issues:

• Paywalled and login-protected pages cannot be scraped. Export them as PDFs first.
• YouTube videos without captions do not provide NotebookLM with readable content.
• Google Docs do not sync automatically. Use the sync button on the source when the underlying document changes.
• Mixing unrelated material in one notebook can make answers less focused.

Features worth using

NotebookLM includes many features, but four are especially practical.

Audio Overviews

Audio Overviews turn your sources into a podcast-style discussion between two AI hosts. Interactive mode lets you interrupt the episode and ask questions. Before generating an overview, you can focus it on a specific chapter or skip background material.

Free users get three Audio Overviews per day.

Video Overviews

Video Overviews create narrated, slide-style explainers from your sources and offer selectable visual styles. They take longer to generate than audio, but work well when you need to explain material to someone who will not read the original documents.

The free tier includes three Video Overviews per day.

Mind maps

Mind maps generate a branching diagram of the concepts across your sources. They are useful for quickly surveying an unfamiliar topic:

1. Generate a mind map.
2. Find a branch you do not understand.
3. Open it and ask a focused question.
4. Follow the citations back to the source.

Reports, flashcards, and quizzes

The Studio panel can turn sources into briefing documents, study guides, flashcards, and quizzes. Students can use these for revision, while teams can generate a briefing document before a meeting on an unfamiliar topic.

These features share a pool of 10 generations per day on the free plan.

In 2026, all tiers use Gemini 3 models. Paying for a higher tier increases usage limits; it does not provide a smarter NotebookLM model.

Free vs. paid NotebookLM tiers

Google offers the upper NotebookLM tiers through Google AI subscriptions rather than as a standalone NotebookLM plan. The main difference is volume:

Limit	Free	Google AI Plus	Google AI Pro	Google AI Ultra
Notebooks	100	200	500	500
Sources per notebook	50	100	300	500–600
Chat queries per day	50	200	500	2,500–5,000
Audio/Video Overviews per day	3	6	20	Up to 200 video

Google AI Pro removes the watermark from generated videos in most regions. Ultra also unlocks the cinematic Video Overview style built on Google’s Veo video model.

Google AI Pro costs $19.99 per month, while the Plus tier has been reported at $4.99. Students in several countries have received Google AI Pro free for a year.

The practical approach is to stay on the free tier until a limit affects your work. Heavy users are most likely to hit the 50-chat-per-day limit first; that is when Plus or Pro may become worthwhile.

Developer workflow: use NotebookLM for API documentation

API documentation is a strong NotebookLM use case: it is long, structured, detail-heavy, and difficult to search when you need one specific answer.

Build an API documentation notebook

1. Create a notebook for the API you are integrating.
2. Add the provider’s reference documentation as URLs or PDFs.
3. Include relevant RFCs, such as OAuth 2.0’s RFC 6749.
4. Add your OpenAPI specification as Markdown or a shareable document.
5. Ask concrete implementation questions.

Useful prompts include:

• “What is the rate limit for the batch endpoint?”
• “Which scopes does the refresh flow require?”
• “Do any endpoints in our specification return a 429 without a Retry-After header?”
• “Which response fields are required for the create operation?”

NotebookLM returns citations to the relevant section of the specification, which is often faster than searching multiple documentation sites. Teams working with Gemini’s own API can use the same workflow to query model parameters and API behavior.

Know what NotebookLM cannot do

NotebookLM reads documents; it does not call API endpoints. It may accurately cite a response schema that your production server stopped honoring several releases ago.

That is where Apidog fits:

• Maintain the OpenAPI specification as a living source of truth.
• Send real requests to the API.
• Validate responses against the schema.
• Detect differences between documentation and production.
• Re-export the validated specification to NotebookLM when it changes.

Use NotebookLM to understand an API, and Apidog to verify that the API still behaves as documented.

FAQ

Does NotebookLM train on my uploaded data?

Google states that NotebookLM does not use sources, queries, or responses from free and paid personal accounts to train its models. Workspace accounts receive additional contractual data protections. Still, apply normal caution when handling sensitive information because policies can change.

Can I share a notebook?

Yes. Notebooks can be shared like Drive files with viewer and editor roles. Viewers can chat with the sources without modifying them. Public links to Audio Overviews are also supported.

Is there a NotebookLM API?

There was no public NotebookLM API as of August 2026. For programmatic source-grounded answers, use Gemini models directly. Gemini 3 Pro through Ollama is an option for local usage, while Google’s File API supports cloud uploads.

Which languages does NotebookLM support?

Chat and generated outputs work in dozens of languages. You can set the output language independently from the source language. Audio Overviews also support many languages, although non-English voices may be less polished than the English hosts.

Conclusion

NotebookLM’s free tier includes the core product: 50 sources per notebook, cited answers, audio and video generation, and mind maps. Start with one notebook focused on a real problem. Once you are comfortable with the workflow, explore these three real-world scenarios.

If your sources are API documentation, complete the loop with Download Apidog. Design the specification, mock it, test it, and keep the documentation aligned with the API that actually runs.