This guide is for people who already generated a few and want output that actually fits a use case. Six adjustments, ordered by how much they change the result.

Why do custom prompts change the video so much?

The generation prompt sets both the script and the visual plan. NotebookLM reads it as the audience brief. Give it a role and a goal, and the model drops the generic survey tone. “Explain the key findings to a busy executive” and “Explain the key findings” are not the same request, and the videos prove it.

Keep the instruction under about 25 words. Longer prompts get partially ignored because the model has to fit them into a fixed runtime.

Six prompt tweaks that work

1. Name the audience. “For a new sales hire who has never seen the product.”
2. Set a hard angle. “Compare our pricing against the two competitors in the source.”
3. Ban a topic. “Do not mention the 2024 pilot, it is outdated.”
4. Force numbers. “Include every dollar figure and percentage from the source.”
5. Set the ending. “Finish with three action items.”
6. Set the tone. “Skeptical, not promotional.”

Combine two or three at most. Four or more and the video starts feeling cramped, because it is trying to satisfy all of them inside the same few minutes.

How many sources should you upload per video?

Five to ten documents is the sweet spot. Below three, the video is thin and repeats itself. Above fifteen, the model spreads attention thinly and the strongest document does not get the screen time it deserves.

Assume one dense source per one to one-and-a-half minutes of video. A 5-minute video comfortably covers 4-6 substantial documents.

Does source order matter?

Yes, more than most people expect. NotebookLM gives early sources more weight in the opening section. Put the document that should frame the whole video first in the source list. If a chart or table is the punchline, upload it as its own source so the model treats it as a first-class item.

When should you regenerate instead of editing?

Regenerating costs you another generation from your daily quota, so it is not free. But there is no sentence-level editor for the narration. If one sentence is wrong, your options are regenerate with a sharper prompt or stop using the video entirely. For anything client-facing, number a single regeneration into your workflow rather than fighting the first output.

Who is this advanced workflow for?

It suits teams publishing internal briefings on a schedule, where a repeatable prompt template saves real time. It is overkill for a one-off summary you will watch once. If you just need the gist of a PDF, the default settings are fine.

Key Takeaways

• A custom generation prompt sets both script and visuals; keep it under about 25 words.
• Five to ten focused sources beat a large dump, and early sources carry more weight.
• Six tweaks — audience, angle, exclusions, numbers, ending, tone — cover most needs.
• Regeneration is the only fix for a bad sentence and it costs a daily generation.
• Worth templating for recurring briefings, overkill for one-off summaries.

That is the 40-word version. The rest of this guide covers what actually happens when you try it on a real 40-page report, where it stumbles, and whether the free tier is enough.

What is NotebookLM Video Overviews?

It is a video generator built on top of NotebookLM, Google’s source-grounded research assistant. Instead of answering your question in text, it produces a shareable video summary. The AI picks the key points across your sources, writes a narration script, and lays it over simple visuals.

Two things matter here. First, everything stays grounded in the files you upload — it should not invent facts outside them. Second, the output is a slide deck with a voice, not a full motion-graphics production. If you expected cinematic B-roll, you will be disappointed.

How do you create a Video Overview step by step?

The workflow is short. The prep before it is what determines quality.

1. Open NotebookLM and create a notebook.
2. Upload your sources: PDFs, Google Docs, pasted text, or a website URL. Keep it to 5-10 focused documents.
3. Wait for indexing to finish (the source chips turn solid).
4. Find the Video Overview option in the Studio panel on the right.
5. Hit Generate. It usually takes 3-8 minutes.
6. Play the result, then use the prompt box to steer a regeneration — for example “focus on the cost model, skip the history.”

You can also type a custom instruction before generating, and that single sentence does more for quality than anything else. “Explain this to a new hire” and “Summarize the risks for a CFO” produce completely different videos from the same file.

What are the real limits?

Free accounts get a limited number of videos per day, and the count resets on a rolling basis rather than at midnight. Long documents get compressed hard — a 60-page source often becomes 8-12 talking points, and nuance disappears first. The host’s tone is pleasant but generic, with the same cadence across notebooks.

There is a smaller gotcha: scanned PDFs without a text layer index poorly or not at all. Run them through OCR first if the video comes out empty.

Who should use it — and who should not?

It fits students condensing lecture readings, analysts who need a quick brief for a meeting, and anyone sharing research with people who will not read a report. A 5-minute video beats a 40-page PDF for that audience every time.

Skip it if you need branded visuals, precise chart reproduction, or a script you fully control. The narration is editable only through regeneration prompts, not word by word. For regulated content where every sentence is reviewed by legal, that is a dealbreaker.

How does it compare to a plain Audio Overview?

Audio Overviews give you two hosts talking. Video Overviews give you one narrator over slides. The video version is easier to share on a landing page or in a Slack channel because people can watch muted with captions. The audio version is faster to generate and better for commutes. Pick based on where your audience will actually consume it.

Key Takeaways

• NotebookLM Video Overviews turns uploaded sources into a narrated slide video in about 3-8 minutes.
• Quality depends on your prompt and source selection more than on any setting.
• Free accounts cap daily video generations; long documents get heavily compressed.
• Best for summarizing research for non-readers; poor fit for branded or legally-reviewed content.
• Scanned PDFs need OCR before upload or the video will be thin.