The new short video format uses Google's Nano Banana 2 Lite model to turn uploaded content into narrated slideshows, but only paid subscribers get meaningful access.
Google just gave its AI notebook tool a TikTok-shaped upgrade. NotebookLM can now generate roughly 60-second vertical videos from user-uploaded notes, turning research materials into bite-sized narrated slideshows complete with summaries, images, and diagrams.
The feature, called Short Video Overviews, is available exclusively to Google AI Pro and Ultra subscribers aged 18 and over. Free-tier users are either locked out entirely or limited to significantly fewer generations.
How the short video format works #
The new Short format builds on NotebookLM’s Video Overviews feature, which first launched on July 29, 2025. Where the original format produced longer-form video summaries, the Short option condenses everything into a vertical, roughly one-minute clip designed for quick consumption.
Under the hood, the feature relies on Google’s Nano Banana 2 Lite model for rapid text-to-image generation, alongside Gemini-family models for the broader content processing. Users upload their notes and sources through the Studio interface, and the system produces a narrated slideshow that pulls key points, visuals, and diagrams from the uploaded content.
The output is English-only for now. Google AI Pro subscribers can generate up to 20 videos per day, while free-tier users face substantially lower caps or no access at all.
Alongside the Short format, Google rolled out additional visual style options and enhanced capabilities using the broader Nano Banana model family. The update also increased generation limits for paid users across the board.
Google’s bigger play in AI content creation #
The original NotebookLM gained traction as a research and note-taking tool powered by AI. Its Audio Overviews feature, which generated podcast-style conversations from documents, became something of a viral hit. Video Overviews represented the natural next step, and the Short format now optimizes that capability for vertical-first consumption.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our