NotebookLM's Free Short Video Overview: Turn Any Article Into A Video (Full Setup Guide)

TL;DR

Brain rot just got named the word of the year, and reading for fun has fallen 40% to its worst level on record. Google’s response, whether they meant it that way or not, is a free NotebookLM feature called Short Video Overview: drop in any article, PDF, or report, and it hands you back a narrated vertical video that teaches the material, complete with a steering prompt so you can point it at exactly what matters and even script your own closing call to action. This guide walks through the setup, the two prompts that make it actually useful, and the research behind why this isn’t rotting your brain further, it’s a second way in that sticks better than text alone.

Why “doom scroll your way to learning” isn’t as crazy as it sounds

Brain rot just got named word of the year, and reading for fun has fallen to its worst level on record. It’s clear social media has rewired our brains and the way we learn. That’s not really a “kids these days” story anymore, it’s an everyone story. We’re wired for quick stories and visuals over walls of text now, and the honest move isn’t to fight that shift with another “just read more” campaign. It’s to meet people where their attention already lives, including the fact that you never actually liked reading 40-page reports anyways.

That’s the gap NotebookLM’s newest update quietly fills.

What NotebookLM’s Short Video Overview actually does

NotebookLM is Google’s free research and note-taking tool, built around the idea that you drop sources in and the tool helps you understand them. It already had a longer-form “Video Overview” feature. The new addition is Short, a vertical, narrated video generated from whatever you feed it, built to be watched the way you’d watch a Reel or a Short, not the way you’d sit through a lecture.

Here’s the actual workflow:

Go to notebooklm.google.com. It’s free, no install, sign in with any Google account.
Create a notebook and add a source, a link, a PDF, a pasted article, a 40-page report, whatever you’ve been meaning to get through.
Click Video Overview in the Studio panel, then select Short instead of the standard length.
Before you generate, type one line telling it what to focus on. This is the step almost nobody uses, and it’s the difference between a generic summary and a video that actually answers the question you had.
Give it a few minutes. It hands back a narrated vertical video with a voice that sounds like an actual person, not a robotic text-to-speech read.

The part worth calling out twice: you can tell it to end the video with your own call to action. That single detail turns a passive summary tool into something you can actually build content or study workflows around.

The steering prompt: the part that changes everything

Most people who try NotebookLM’s video features stop at “generate” and take whatever comes out. That’s leaving the best part on the table. The steering box lets you narrow the output before it ever renders, which matters because a 40-page report and a 40-page report summarized for “the three biggest financial risks” are two completely different videos.

A simple template that works across most source types:

Focus on [the main argument / the 3 biggest takeaways / the part I keep forgetting].
Keep it simple and use plain language, no jargon.

Swap the bracketed part for whatever you actually need out of the source. Studying for an exam, focus on definitions and dates. Reading a competitor’s earnings call, focus on guidance changes. Trying to finally get through that report your team keeps referencing, focus on the recommendation and why.

The CTA-insertion prompt

This is the detail that separates a study tool from a content or teaching tool. In the same steering box, you can tell NotebookLM exactly how to close the video:

End the video with this exact call to action, spoken naturally:
"[your CTA here, e.g. save this and try it yourself]"

That means the short doesn’t just summarize, it can direct. If you’re using this to teach a concept to your team, end it with “try this yourself with your own report.” If you’re using it to process your own reading list, end it with a reminder of what to do with the takeaway. Small detail, real difference in how usable the output ends up being.

Run it twice

It’s free, so there’s no reason to settle for the first pass. Generate once with a broad summary steering prompt. Generate again with a narrower angle, a specific section, a counterargument, the practical takeaway. Keep whichever version actually lands for what you needed, and throw out the other. Two free generations beats one mediocre one every time.

This isn’t rotting your brain further, here’s the research

The obvious objection: doesn’t watching a video instead of reading the source material just add to the doom scroll problem in the first place? The research says the opposite, at least when the video and the source work together instead of replacing each other outright. You were never going to read that 40-page report. You’ll watch a 60-second video. That’s not brain rot, that’s hijacking the format that’s been frying everyone else’s attention.

Richard Mayer at UC Santa Barbara, who’s spent decades studying multimedia learning, found that people score a median of 89% higher on transfer tests (tests that measure whether you can apply what you learned, not just recall it) when they learn from words and visuals together compared to words alone. That’s not a rounding-error improvement. That’s the multimedia learning effect doing exactly what four decades of cognitive science says it should do: pairing narration with visuals gives your brain two channels to process and store the same information instead of one.

So the short video isn’t a shortcut around the reading. It’s a second entry point that makes the material stick better than the article would have on its own, especially for material you were realistically never going to sit down and read cover to cover anyway.

Who this is actually for

You don’t need to be a teacher or a student to get value out of this. Anyone with a stack of tabs they’ve been meaning to read, a report they keep meaning to get through, or a topic they need to understand quickly has a use for this. Founders processing competitor research. Anyone catching up on an industry report before a meeting. Parents trying to actually finish the article they saved three weeks ago. The use case is the same in every version: turn the thing you were never going to read into the thing you’ll actually watch.

FAQ

Is NotebookLM’s Short Video Overview actually free? Yes. It launched free to all NotebookLM users, no Pro or Ultra subscription required, no separate payment.

How long does it take to generate a short? Typically five to ten minutes from clicking generate to having a finished vertical video.

Can I use it on a PDF, not just a web link? Yes. NotebookLM accepts links, uploaded PDFs, and pasted text as sources, and the Short Video Overview works from any of them.

Does the narration sound robotic? No, the narration is designed to sound like a real person walking you through the material, not a flat text-to-speech read.

Can I control what the video focuses on? Yes, and this is the step most people skip. Type a one-line steering prompt before generating to tell it exactly what to emphasize, and you can even script the closing call to action.

Try it on the article you’ve been meaning to read

Pick the one link or PDF you’ve had open in a tab for two weeks. Drop it into NotebookLM, hit Video Overview, select Short, and use the steering prompt above. Five minutes from now you’ll have watched the thing you were never going to read.

If you want the full copy-paste setup, including both prompt templates from this guide, comment “SHORTS” on the video and it’ll land in your DMs. And if you want a new setup like this one delivered every week, the Actionable AI community is where those live first.