NotebookLM May Soon Create Video Summaries

Google seems to be working on an AI-generated Video Overview feature for its NotebookLM. A new hidden feature has appeared on the Illuminate, which is an experimental project. This feature is expected to follow the popular Audio Overviews feature launched last year.

While Google hasn’t officially confirmed the rollout, the company did hint at big things coming through its Google Labs account on X (formerly Twitter), saying:

That “fun stuff” may include short, fully AI-generated videos created from documents and questions. This feature is already quietly being tested inside Google. 

What is Sparks? 

According to the report from TestingCatalog, the experimental Illuminate project now shows a new section called Sparks, marked under the “Early Preview.” It says, “Imagine any question could be instantly transformed into a short video, 100% AI-generated.”

These vertical videos are one to three minutes long and appear to be created using a single model that generates synchronized audio and visuals. While the feature is currently limited to internal Google accounts, it suggests a future where you could upload a document and instantly receive a video summary hosted by AI characters. 

NotebookLM Could Be Next 

The overall format of videos closely aligns with NotebookLM’s upcoming Video Overviews feature, where two AI hosts present conversational video summaries based on uploaded content. The shared goal of summarising complex materials into simple outputs makes it very likely that the underlying tech stack is connected, possibly even powered by Veo 3 or a multi-modal Gemini Ultra variant.

Only time will tell whether these Video Overviews arrive on NotebookLM right at Google I/O or shortly after. However, one thing is certain: Google is sprinting toward AI-powered, end-to-end media generation. The company recently announced that NotebookLM’s mobile apps will be available for Android and iOS devices. This means that if the Video Overview feature makes it to NotebookLM, users will be able to create AI-generated short vertical videos from their smartphones.