Google Gemini Introduces Agentic Video Understanding to Reduce Costs and Boost Accuracy
Google has rolled out "agentic video understanding" across its Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite models. This new feature significantly cuts costs of video analysis by up to 66% and reduces token usage by up to 88%, while boosting accuracy. It uses Gemini's reasoning combined with video tools to dynamically process content, allowing for specific moment seeking, anomaly detection, and precise counting.
Small business owners can now more affordably and accurately analyze video content, for example, to understand customer engagement with marketing videos or track employee interactions with training materials. This provides actionable insights to refine strategies and improve efficiency.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free