Google DeepMind Adds Agentic Video Understanding to Gemini Models

Google DeepMind launched agentic video understanding for Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite. The feature lets models dynamically search, scan, and inspect video segments across frames, audio, and transcripts, rather than ingesting at a fixed frame rate. According to Google DeepMind, this reduces token consumption by up to 88% and costs by up to 66%, while improving accuracy by up to 7% on standard benchmarks. Gemini 3.7 Flash with the feature sits at the accuracy-to-cost Pareto frontier among the models tested. Newly enabled capabilities include sub-second moment retrieval, long-form needle-in-a-haystack search, anomaly detection, and accurate action and object counting. The feature will power YouTube's 'Ask YouTube' on the video watch page and will roll out to Gemini app users across Flash and Flash-Lite models.