Google Gemini Unveils Agentic Video Understanding to Cut Costs and Boost Accuracy

Google has launched 'agentic video understanding' within its Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite models. This new feature significantly reduces token usage by up to 88% and analysis costs by as much as 66%. It allows for dynamic processing of video content, leveraging Gemini's reasoning to seek specific moments, detect anomalies, and count elements within videos, rather than relying on static frame-by-frame analysis.