Gemini gains agentic video understanding for YouTube's Ask

Google introduced agentic video understanding across Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite, letting the model dynamically scan and choose which video segments to analyze instead of processing fixed frame intervals. Per Google, the approach cuts video token consumption by up to 88%, lowers cost by up to 66%, and improves accuracy by up to 7%, while adding sub-second moment retrieval and long-form video search. The feature is live now via the Gemini API in Google AI Studio and the Gemini Enterprise Agent Platform, rolling out to the Gemini app soon, and Google says it will power YouTube’s “Ask” feature on video watch pages in the coming months.