Google launches agentic video understanding in Gemini, using dynamic FPS to count fast actions like claps

Google on September 1, 2026 launched agentic video understanding across Gemini 3.7 Flash, Gemini 3.6 Flash, and Gemini 3.5 Flash-Lite, a processing mode that lets the model decide what to watch, at what speed, and through which modality instead of ingesting video at a fixed frame rate. In a demonstration, Gemini 3.7 Flash counted rapid claps by automatically adapting processing speed—work that static analysis often misses because it samples video at a default of 1 frame per second (FPS).