# Google's Gemini Omni 1.2 Flash video model goes GA with 15-second clips

published: 2026-11-05T23:05:54.696Z
topic: infrastructure
source: https://the-decoder.com/googles-gemini-omni-1-2-flash-stretches-ai-video-clips-to-15-seconds-and-cuts-pricing-by-30-percent/
url: https://aiagentnews.io/stories/google-s-gemini-omni-1-2-flash-video-model-goes-ga-with-15-second-clips

Google DeepMind has made its Gemini Omni 1.2 Flash video model generally available, with clips up to 15 seconds long. The update adds conversational editing and cuts pricing by 30 percent to $0.07 per second for 720p video.

Google DeepMind is making Gemini Omni 1.2 Flash, its video-generation model, generally available. Clip length now runs from three to 15 seconds, up from a previous maximum of 10 seconds. Google says the model is rolling out through the Gemini API, Google AI Studio, Google Cloud Console, Flow, Google Vids, the Gemini app for Google AI Ultra subscribers and Google Ads.

A feature Google calls conversational editing lets users change a generated clip step by step, such as its background, lighting or aspect ratio, without regenerating the whole clip. Google says this works across the full 15-second length. It also proposes a tiered workflow: drafts at 360p or 720p, detail fixes through follow-up edits, and upscaling to 1080p or 4K only for the final result.

Google says the model accepts character images, product photos, motion clips of up to three seconds and audio cues as inputs, and aims to keep faces, clothing and brand colors consistent across camera angles. It also reports fewer visual drifts and jump cuts, along with improved lip-sync and handling of multiple speakers. Every generated video carries an imperceptible SynthID watermark.

## Why it matters if you build agents

Developers calling the Gemini API for video can now request clips up to 15 seconds, priced at $0.07 per second for 720p after the 30 percent cut. Editing a draft through follow-up changes instead of regenerating it may reduce iteration costs, though the savings depend on how a pipeline is built. Teams that process generated video should also account for the SynthID watermark in their outputs.