Introducing Agentic Video Understanding with Gemini
Gemini models can now scan video dynamically, jumping between segments instead of processing every frame at a fixed rate, which Google says cuts token use…
Gemini can now watch video the way a person skims it: jumping to the relevant segments instead of processing every frame at a fixed rate. The new agentic video understanding feature lets Gemini models decide what to watch, at what speed, and through which modality (frames, audio or transcript), fetching only the moments needed to answer a query.
Google says the approach cuts token consumption by up to 88% and analysis costs by up to 66%, while improving accuracy by up to 7%, with the biggest gains on long-form video from 10-minute how-to guides to multi-hour recordings. It enables things like sub-second moment retrieval, needle-in-a-haystack search across hours of footage, anomaly detection at higher frame rates, and accurate counting of repeated actions or objects.
The feature is live now on Gemini 3.7 Flash, 3.6 Flash and 3.5 Flash-Lite, for both uploaded video and YouTube links, via the Gemini API in Google AI Studio and the Gemini Enterprise Agent Platform. It uses standard token pricing with no extra fee; developers turn it on by setting processing to "agentic" in the API configuration. Google plans to bring it to the Gemini app for Flash and Flash-Lite models, and to YouTube's Ask YouTube feature, in the coming months.
