Skip to content
Trending storyDeveloping

Google adds agentic video understanding to Gemini API, citing up to 66% lower cost

1 article1 sourcesince Oct 19Last article 1h ago ·

Overview

AISummary of 1 article

Google for Developers says agentic video understanding in the Gemini API fetches only the video slices a prompt needs through tool calls.

The company reports up to 88% fewer tokens, up to 66% lower cost, and up to 7% higher accuracy on long-form video compared with 1 FPS static sampling.

Developers turn it on by setting processing to "agentic" in the request. The company says it is supported on Gemini 3.8 Flash, 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite, and can be used with the Gemini Batch API at 50% of standard cost.

Written by AI from the articles below · updated Oct 19, 1:58 PM ET

Check the sources:

Latest developmentOct 19, 1:45 PM ET
Google adds agentic video understanding to the Gemini API for long-form video

Article timeline

The articles in this story. Times are ET.

Oct 19
  1. Google for DevelopersOfficial
    Google adds agentic video understanding to the Gemini API for long-form video

    AIGoogle for Developers says agentic video understanding in the Gemini API uses tool calls to fetch only the video slices a prompt needs. The source reports up to 88% fewer tokens, up to 66% lower cost, and up to 7% higher accuracy on long-form media versus 1 FPS static sampling. Developers enable it by setting processing to "agentic" and can use it with the Batch API at 50% of standard cost.

Heat Trend

Not enough continuous observations to show a trend yet.