NotebookLM Flashcards & Quizzes — what shipped

NotebookLM's Flashcards and Quizzes features ship on web and mobile, free and Workspace alike, generating study cards grounded only in your uploaded source materials. The 2026-04 update added persistent session memory (resume where you left off), shuffle, delete, and mastery tracking — bringing the feature closer to mature tools like Anki and Quizlet. Customize topics, difficulty, number of questions, and the source subset. This page recaps what shipped and how BibiGPT's transcript-grounded summary + Q&A pipeline covers the same study workflow for video and podcast content.

Launched · 2025-09 Updated · 2026-04 Web + mobile · free + Workspace

Key facts (90-second read)

NotebookLM's Flashcards and Quizzes shipped on web and mobile, free and Workspace alike, generating study cards strictly grounded in your uploaded sources (PDFs, slides, YouTube videos, websites, Google Docs). The 2026-04 update added persistent session memory, shuffle, delete and mastery tracking — closing the daily-study gap with Anki and Quizlet. Customize topic, difficulty, audience style and number of questions per generation. For video and podcast study, BibiGPT's transcript-grounded summary + Q&A pipeline covers the same source-grounded workflow across YouTube, Bilibili, podcasts and uploaded files, with 5-language output (zh / en / ja / ko / zh-tw) and timestamped jumps.

Features

What Flashcards & Quizzes ship today

Source-grounded study cards generated in seconds from your uploaded materials. The feature works across web and the mobile app, free tier and Workspace, with no distinction.

Generated only from your sources

Both Flashcards and Quizzes are grounded strictly in the source materials you upload — PDFs, slides, YouTube videos, websites and Google Docs. NotebookLM does not pull facts from the wider web, which makes the cards safer to study from for course-level material.

Customize topic, difficulty, count

Add a prompt to focus on a specific topic, difficulty level, or audience style. Adjust the number of questions, pick a source subset, and regenerate variants when you want fresh practice questions on the same material.

Persistent session memory (April 2026 update)

The 2026-04 update added persistent progress: leave a flashcard deck and pick up where you left off, shuffle the deck, delete cards you've mastered, and track what's been mastered vs what needs review. Closes the gap with Anki / Quizlet for everyday study loops.

Where Flashcards & Quizzes work best

Three concrete scenarios where NotebookLM's source-grounded study cards genuinely save time.

Reading-heavy coursework

Upload course PDFs and slides into one notebook, generate flashcards per chapter, and run a quiz before exams. Source-grounded means the answers always trace back to your actual reading list.

Self-paced certifications

Upload official study guides and changelog pages. Generate question sets at increasing difficulty levels for spaced practice over weeks, with mastery tracking to focus on the weak spots.

Quick comprehension checks on long sources

Have a 200-page report or a 3-hour lecture series? Generate a 10-question quiz to validate your understanding before going deeper, then create flashcards on the specific concepts you got wrong.

How BibiGPT compares for video and podcast study

BibiGPT's transcript-grounded summary + Q&A pipeline covers the same source-grounded study workflow for video and podcast content — across YouTube, Bilibili, podcasts, uploaded files and lectures.

Transcript-grounded notes from any video URL

Paste a YouTube, Bilibili or podcast URL. BibiGPT extracts the transcript, generates a multi-section summary grounded in the spoken content with timestamped jumps, and lets you ask follow-up questions in the same conversation.

5-language output for global study groups

BibiGPT generates the notes in zh / en / ja / ko / zh-tw. Useful when your study group splits across language regions — the same source video produces notes and Q&A in each member's working language.

Mind maps + AI chat for deeper comprehension

Beyond bullet-list summaries, BibiGPT auto-generates a mind map of the video structure and keeps the transcript indexed for follow-up Q&A. Pair with NotebookLM Flashcards for the recall-practice layer if you want both — they complement rather than overlap.

5 key facts (90-second read)

What shipped in NotebookLM Flashcards & Quizzes and what the 2026-04 update changed.

  1. 1

    Available on web + mobile, free + Workspace

    Both Flashcards and Quizzes work on the web app and mobile app, with no feature gating between free-tier and Workspace users. No paid wall, no enterprise-only restrictions.

  2. 2

    Strictly grounded in your sources

    Cards and questions are generated only from the source materials you upload — PDFs, slides, YouTube videos, websites and Google Docs. NotebookLM does not pull facts from the wider web for these features, which keeps the study material safe for course-level work.

  3. 3

    Customize topic / difficulty / count / sources

    Add a prompt to focus on a specific topic, difficulty level, or audience style. Adjust the number of questions, choose a source subset, and regenerate fresh question variants on the same material.

  4. 4

    2026-04 update adds persistent memory + mastery

    Resume sessions where you left off. Shuffle the deck. Delete mastered cards. Track which cards you've mastered vs which need review. Closes the daily-recall gap with Anki and Quizlet.

  5. 5

    BibiGPT pairs as the video-source layer

    For YouTube, Bilibili, podcast and uploaded video sources, BibiGPT generates transcript-grounded summaries, mind maps and timestamped Q&A in 5 languages. Pair with NotebookLM Flashcards for the recall-practice layer — they complement rather than overlap.

3 typical scenarios for BibiGPT + NotebookLM users

Where the two tools combine cleanly for end-to-end study workflows.

Lecture video → transcript notes → flashcards

Paste a YouTube university lecture into BibiGPT, get a transcript-grounded summary with timestamped jumps and a mind map. Export the summary as text and feed it into NotebookLM as a source to generate flashcards for recall practice.

Podcast episode series → cross-episode quiz

Use BibiGPT to extract transcripts and per-episode summaries from a podcast series. Drop the transcripts into a NotebookLM notebook and generate a quiz that spans the whole series — testing comprehension across episodes.

Multilingual study group

BibiGPT outputs summaries in zh / en / ja / ko / zh-tw for the same source video. Each study group member gets notes in their working language; NotebookLM Flashcards can then be generated per-language from those translated summaries for shared recall drills.

Frequently Asked Questions

Ask us anything!

Generate transcript-grounded notes from any video URL with BibiGPT — pair with NotebookLM Flashcards for recall practice

Paste a YouTube, Bilibili or podcast URL into BibiGPT. Get a multi-section summary, timestamped jumps, mind map and follow-up Q&A in zh / en / ja / ko / zh-tw — all grounded in the actual transcript. Then use NotebookLM Flashcards on top if you want spaced-recall drills.