r/notebooklm

Suddenly the Pro badge disappeared from my NotebookLM, now it's downgraded back to free tier even tho I have Google AI Pro... 🤷‍♂️

Is anyone having an issue with their account? The reason I’m asking is because currently my pro account is being seen as a free account because I’m seeing beta for slide deck and infographic. Also, Gemini is arguing me down that there is an issue that I am something I am doing wrong.

My account says Pro but my sources are limited to 50. I get my Google Pro via a Mobile add-on, am I experiencing a Pro(minus)?

Quick heads up for fellow NotebookLM folks: there’s an active Google Workspace Status Dashboard notice right now about some paid-tier accounts showing Free tier features/limits.

Not affiliated, just sharing so nobody spends 45 minutes gaslighting themselves. 😅

Status link:
https://www.google.com/appsstatus/dashboard/summary

https://preview.redd.it/1nnd47qb372h1.png?width=1254&format=png&auto=webp&s=af3a3454d0271330f1380669e24f933030c23904

I stopped treating NotebookLM as just a summarizer. Lately I've been using it as a second stage research workspace. My workflow looks like this

1 Collect source material

- articles

- PDFs

- transcripts

- my own notes

2 Use NotebookLM to explore

- ask clarifying questions

- compare sources

- identify contradictions

- generate timelines and mind maps

3 Extract what is actually worth keeping

- key observations

- claims that need verification

- open questions

- practical takeaways

What changed for me is that NotebookLM works best when you already have a focused question. If I dump sources in and ask for "a summary" the output is useful but generic. If I ask

- "What assumptions are repeated across these sources?"

- "Where do these sources disagree?"

- "Which claims are based on evidence vs opinion?"

- "What would be risky to state publicly without verification?"

The results become much more valuable. One unexpected benefit - it helps separate learning from content creation. First I use NotebookLM to understand the topic. Only after that do I turn the findings into a post, article, or research note. That small separation has made my writing more precise and much less hype driven.

Curious how others are using NotebookLM. Do you mostly use it for summarization, or has it become part of a larger research workflow?

I’ve been deep in NotebookLM for the last few months, and the May 2026 update just made Mind Maps feel like an entirely new tool.

Until recently, you’d generate a mind map and just accept whatever structure the AI decided to give you. Now? You can actually tell it exactly what kind of map you want, i.e., the hierarchy, the logic, colors, cross-connections, or even the type of thinking you want it to do (see my prompt below and mind map generated in response as the image).

This isn’t just a small tweak. It’s the difference between “pretty summary bubbles” and actually useful thinking structures.

The 6 patterns I’m using the most right now:

1. Tension Map — My favorite by far. Instead of a boring central topic, you put a contested claim in the middle (e.g. “AI safety risks are mostly near-term, not existential”). It then builds out steel-manned positions for and against, meta views, and — best part — highlights the real tensions and contradictions between sources. Perfect for literature reviews and dissertation work.
2. Decision Architecture Map — Excellent for strategic choices (build vs buy, hire vs agency, go-to-market options, etc.). It clearly shows options, tradeoffs, risks, and even flags a recommended path with evidence.
3. Literature Constellation — When you have 20–40 papers and don’t want to force a fake synthesis. It creates navigable clusters so you can actually explore the field.
4. Content Ecosystem Map — Turns one big idea into a full multi-channel plan (blog → newsletter → podcast → LinkedIn → YouTube).
5. Second Brain Navigation Map — Life-saver once your NotebookLM vault hits 30–50 sources.
6. Meeting Synthesis Map — Pulls patterns across a bunch of sales calls, user interviews, or team meetings.

Pro tips that made a big difference:

• Always phrase the central node as a full sentence or question, never just a topic.
• Add the line: “Every node must have a citation. If you can’t cite it, don’t include it.” — This keeps everything grounded.
• Paid users (Pro/Workspace/AI Premium) get the native customization panel. Free users can paste the full prompt into the chat and then tap the Mind Map chip — it works really well after 1-2 iterations. I tested in both my paid account and my free account and was able to get similar results in my free account with a few attempts telling NotebookLM to generate mind map based on my prompts.

I put together a free PDF with all 6 full prompts. More resources in my profile.

Below is the prompt I used for the image  shared here and I am pretty happy with the end product. What’s even better is how one click on any of the nodes starts a full search session before NotebookLM quickly generates solid content about the topic.

Prompt: You are a conceptual cartographer. Create a Mind Map revealing non-obvious cross-source connections for GEO/AEO. Use extensive cross-links between branches. Group under categories like Assumptions, Challenges, Dependencies, Implications, Opportunities. Add novel synthesis nodes explicitly labeled 'AI-Derived Insight: [brief description]' with grounding

If you do any kind of research, writing, content creation, or strategic thinking, this update is honestly worth checking out. The citation anchoring is still NotebookLM’s biggest advantage over Whimsical, Miro, or MarkMap.

Who else is playing with the new steering? What kind of map are you building right now? Would love to hear your workflows!

When I upload a source (like a PDF I created), does NotebookLM use the information from the PDF to train its AI?

I'm completely clueless about this..

Is there a way to create folders in NotebookLM? I use it for a wide range of things .. some personal, some professional, some travel-related ... and after a while, it becomes difficult to organize and navigate everything.

Is there any way to create folders in NotebookLM?

We hired two new engineers this year. Every time someone joins I end up spending half a day walking through the same stuff
and created solution with

NotebookLM
Distilbook

Spent a weekend throwing everything into a NotebookLM notebook.

All our internal docs, some YouTube videos I'd saved, confluence pages, whatever. Asked it to index everything into topics first before doing anything else. Then went one topic at a time

For the conceptual stuff
audio overviews. Short ones. New hire listens on day one while their environment is setting up.

Video Making:
after different reseruce i ran thorough DistilBook, it makes an actual animated walkthrough of the visuals
I do use notebookllm video overview for short vidoes ..

After their first week I send them a quiz generated from the same notebook. NotebookLM does this in Studio. Scenario-based questions, .

Whatever they get wrong that's the only thing I actually need to sit down and explain.

That's it. Notebook stays, grows, next hire gets the same thing already updated.

Curious if anyone else is using it for team stuff or mostly personal learning here.

After the new update, the mind map feature appears to generate the map, but nothing is actually shown on the screen. The loading time is unusually long, and in some cases the map never loads at all.

This feels like a rendering or display issue rather than a generation issue, because the feature seems to complete in the background but fails to present the result properly

I’ve been using NotebookLM for studying, and I really like the general idea of uploading materials and being able to chat with them. For understanding topics, getting summaries, or asking questions about a document, it can be very useful.

But when I try to use it specifically for exam preparation, I keep feeling like something is missing.

The biggest issue for me is that it helps with understanding, but it does not really turn the material into a structured study workflow. I still have to decide what to review, when to review it, and how to make sure I am actively recalling the content instead of just reading summaries.

Some problems I keep noticing:

• It is good for understanding material, but not as good for memorising it.
• It does not feel very focused on spaced repetition or exam deadlines.
• I still need another tool if I want proper flashcards or active recall.
• The generated summaries are useful, but they do not always map cleanly to what I actually need to practice.
• It feels more like a research/chat tool than a full exam revision tool.

I’ve been looking at alternatives that are more focused on active recall and review sessions. For example:

• Anki is very powerful for spaced repetition, but it takes a lot of manual setup and does not feel very connected to uploaded lecture materials or exam dates.
• Quizlet is easier and more polished, but it feels more like simple flashcard practice than a complete exam preparation workflow.
• RemNote is good if you want notes and flashcards together, although it can feel like a bigger system you have to fully commit to.
• Knowt is useful for turning notes into study materials, but the workflow still feels a bit separate from actual exam planning.
• StudyFetch / Gizmo / similar AI study tools are interesting, but I sometimes find them either too generic or too focused on quick generation rather than managing revision over time.
• ExamFlow is another one I came across / have been trying, which is more focused on uploaded course material, generated flashcards, and exam-date-aware spaced repetition: https://examflowapp.com

I guess what I am looking for is something closer to: “I have an exam on this date, help me study efficiently until then,” instead of only “summarize this PDF.”

Curious if others feel the same way about NotebookLM. Do you mainly use it for understanding and summaries, or have you found a good workflow for actual exam revision and memorisation?

Also interested in hearing what alternatives people are using for AI-assisted studying.

NotebookLM is genuinely a great tool. I use it. But the podcast feature is fundamentally broken right now, and looking at the sub, I'm not the only one frustrated. Inconsistent generation, hangs, and even when it does run cleanly the output often doesn't hold up.

My primary use case for it was the same as a lot of people here: consume academic literature without losing a Saturday to it.

So I ended up putting together two things to scratch my own itch. Zero strings attached.

1. PaperCast. Weekly rotating podcasts across 7 disciplines. Papers are picked based on gravity and momentum (engagement signals plus recent traction), so it surfaces what's actually moving in each field rather than just what's recent. Each episode is a 15-min conversational walkthrough of a single paper, based on the actual paper content rather than a paraphrase. First episode is available without signup.

2. Debrief. RAG for academic papers. I loved what NotebookLM did conceptually but kept running into the hallucination problem, especially on dense technical claims. Debrief is built for that specific use case. Semantic search across a corpus, answers come with clickable references back to the source paper, and the system doesn't try to stuff whole documents into context (which is where a lot of the hallucination happens in general-purpose RAG).

Both live at SOTA Institute (link in profile). Nothing monetarily gated, no upsell.

What I actually want from this post is feedback on the shape:

• For folks who've felt the NotebookLM podcast feature breaking: what specifically is failing for you? Is it the generation, the audio quality, the output not matching the source, or something else?
• For the RAG side: where does NotebookLM (or whatever you're using) actually break for academic work?
• What's missing that would make a tool like this genuinely useful for the way you read papers?

We've all used NotebookLM and we all see the value. But a general-purpose tool isn't always the right shape for a specific use cases.

Hey everyone,

So, we all know that NotebookLM is amazing for research, but once i started using it heavily, i kept running into annoying workflow problems:

• Saving sources from YouTube, Reddit, AI chats, PDFs, etc. was messy
• Sometimes I accidentally saved sources into the wrong notebook
• Organising large research projects became painful
• Repetitive prompts
• Weak export options
• Annoying deletion of multiple notebooks.

I kept building random workarounds for myself and eventually built a Chrome extension to solve some of these problems.

It started as a simple capture tool, but v1.3.0 just got approved today, and it now lets you:

• Saving content directly into NotebookLM in one click
• Import sources from a different notebook to the current one
• Organising sources into collections
• Create and use reusable prompts.
• Export to HTML, Markdown, or PDF.
• Syncing across devices

It started as a personal tool and recently became something I’m sharing with others.

I’m genuinely curious:

What’s your biggest frustration with NotebookLM right now?

And if anyone here uses NotebookLM heavily, I’d be happy to give a few people free 3-month pro access in exchange for honest product feedback.

Thanks

hey lads, i have a problem with notebooklm’s presentation mode.
ive try to make a presentation using notebooklm and everytime i try it says „failed to generate. try again later”
what might be wrong??