gemini Gemini Blog ·

Gemini Introduces Agentic Video Understanding for Enhanced Analysis

aigaengineer
feature announcement

Google is launching agentic video understanding across its Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite models today. This capability enhances video analysis by dynamically searching and inspecting video segments, reducing token consumption by up to 88% and costs by 66% while improving accuracy by 7%. It enables advanced features like sub-second moment retrieval and precise anomaly detection for long-form content. Developers can access this feature immediately via the Gemini API, with wider integration into Google products coming soon.

Features (1)
  • Agentic Video Understanding for Gemini Models

    Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite now support agentic video understanding, enhancing video analysis by dynamically searching and inspecting target segments across visual frames, audio, and transcripts. This reduces analysis costs by up to 66% and token consumption by 88% while improving accuracy by 7%, especially for long-form videos. It unlocks capabilities like sub-second moment retrieval, accurate anomaly detection, and complex queries, and is available today via the Gemini API in Google AI Studio and the Gemini Enterprise Agent Platform.

Read the original announcement →

https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-agentic-video-in-gemini/

Related releases