Google Launches Gemini Omni 1.2 Flash With 15-Second Video Generation
Google DeepMind has introduced Gemini Omni 1.2 Flash, a major generative video model offering native 15-second video generation and multi-turn conversational editing capabilities. The release expands maximum single-generation video duration from 10 seconds supported in Gemini Omni 1.1 to 15 seconds, representing a 50% increase in continuous scene length. Google established pricing for standard 720p video generation at $0.07 per second, delivering a 30% cost reduction compared to rates charged under Gemini Omni 1.1. Developers can access the model through the Gemini API and Google AI Studio, while enterprise organizations can deploy it via the Google Cloud console.
Clip Duration Expansion and Google Workspace Integration
Alongside developer access, Google integrated the new video generation engine directly into Google Vids, its productivity video creation tool for Google Workspace users. Company announcements indicate that individual account holders and corporate Workspace subscribers can produce up to three 15-second video clips monthly within Vids without additional subscription charges. Creators can specify video durations in one-second increments ranging from three seconds to 15 seconds, drafting initial content at 360p or 720p resolutions. The model supports conversational video editing, allowing users to modify specific elements such as swapping backgrounds, adjusting lighting, or altering aspect ratios without generating an entirely new video from scratch.
Multi-Turn Conversational Editing and Brand Consistency
Gemini Omni 1.2 Flash improves character and brand consistency across diverse camera angles, maintaining facial features, garments, packaging typography, and corporate color palettes. Creators can guide video generation by supplying reference character images, product photos, style cues, audio prompts, or short motion clips up to three seconds long. The model supports sophisticated camera choreography, including first-person view flythroughs, orbital panning shots, and wide zoom-outs. Audio generation features have been refined to produce natural dialogue cadence, enhanced multi-speaker separation, and synchronized lip-sync matching on-screen characters. All video outputs from Gemini Omni 1.2 Flash carry an imperceptible SynthID watermark to support media provenance verification and deepfake detection, according to Google.
Commercial Pricing Reductions and Upscaling Pipelines
These generative video and audio enhancements are designed to power marketing workflows across Flow, Google Vids, and Gemini for business. Users who generate initial draft clips can subsequently upscale finalized video outputs to high-definition 1080p or 4K resolution tiers. By lowering API processing fees to $0.07 per second, Google is leveraging competitive infrastructure pricing to challenge alternative AI video platforms in enterprise creative production. Integrating generative video into office productivity suites allows organizations to produce video assets without navigating specialized standalone developer environments, according to Tech-insider. The expanded 15-second duration enables corporate communication teams to assemble cohesive social media marketing assets and digital presentations within existing document workflows. Google Cloud made Provisioned Throughput available for enterprise Gemini Omni 1.2 Flash workloads to support scalable, reliable production rendering.