Google Releases Gemini Omni 1.2 Flash Video Model

Google DeepMind is rolling out Gemini Omni 1.2 Flash, making the video generation model generally available across the Gemini API and Gemini for business. The update introduces native video creation up to 15 seconds, multi-turn conversational editing, and enhanced reference adherence for business and developer workflows.

Google says pricing for Gemini Omni 1.2 Flash is 30% lower than Gemini Omni 1.1, set at $0.07 per second for 720p video. Workflows can start with fast drafting at 360p or 720p, apply follow-up edits to specific elements without re-generating full clips, and upscale final creations to 1080p or 4K.

Clips can be generated and edited natively up to 15 seconds long, an increase from 10 seconds in Omni 1.1. Duration can be specified in one-second increments between 3 and 15 seconds.

Gemini Omni 1.2 Flash allows users to combine character, product, or style images with audio cues and short motion clips up to three seconds. The model maintains consistency for faces, garments, packaging typography, and brand colors across camera angles and shot transitions. Multi-turn conversational editing lets users swap backgrounds, adjust lighting or weather, or change aspect ratios across 15-second clips step by step.

Additional features include multi-speaker conditioning with lip-sync, dialogue generation, and camera controls for FPV flythroughs, zoom-outs, and orbits. Outputs carry an imperceptible SynthID watermark, and Provisioned Throughput is available. The model is accessible today across the Gemini API, Google AI Studio, the Google Cloud console, Flow, and Google Vids.