Google is releasing Gemini Omni 1.2 Flash, now generally available through the Gemini API and Gemini for business. The update adds native 15-second video generation, multi-turn conversational editing, improved adherence to visual and audio references, and cinematic control over camera movement, while lowering the price for 720p video to $0.07 per second (a 30% reduction from Omni 1.1).
What’s new
- Longer native clips: Omni 1.2 Flash can natively generate and edit videos from 3 to 15 seconds, specified in one-second increments. Previous Omni 1.1 supported up to 10 seconds.
- Faster, more cost-effective workflows: creators can draft at 360p or 720p, make follow-up edits to fix details rather than re-generating the whole clip, and upscale selected results to 1080p or 4K.
- Price: 720p video is now priced at $0.07 per second, a 30% decrease versus Gemini Omni 1.1.
Improved reference fidelity and consistency
Gemini Omni 1.2 Flash better preserves details across frames and camera angles: faces, garments, packaging typography and brand colors remain consistent. The model accepts combinations of character, product or style images, short motion clips (up to three seconds), and audio cues and maintains those references throughout generation and edits.
Examples provided by Google include:
- Combining a model photo with front-and-back product shots of multiple outfits to generate a 15-second virtual try-on sequence that preserves the model’s identity and each garment’s cut, drape and back details.
- Building a layered poster scene sequentially so that materials, shadows and typographic elements remain locked consistently through the final frame.
Motion, camera control and cinematic look
The model supports more complex continuous camera movements (FPV flythroughs, dramatic zoom-outs, orbits) with fewer unwanted jump cuts, steadier scaling, and a polished cinematic aesthetic.
Multi-turn conversational editing
Users can iteratively refine the same clip in conversation-style turns — swapping backgrounds or products, altering lighting or weather, or changing aspect ratio — without starting from scratch. This makes producing variant ads and placements easier without re-briefing or full re-generation.
More natural speech and multi-speaker conditioning
Omni 1.2 Flash generates more fluid, natural-sounding dialogue with clearer lip-sync and better speaker routing for multi-character scenes, which benefits multi-character directing in Flow, explainers in Gemini for business, or polished avatar-driven videos in Google Vids.
Illustrative capabilities cited by Google
- 15-second long clips that keep characters, text and camera moves consistent across multiple turns and transitions with no cuts.
- Multi-turn motion transfer: an initial animated motion (e.g., a rippling octopus) can be replaced in a later turn by another subject (e.g., a dancer) that mirrors the same fluid motion.
- First-to-last-frame consistency improvements over Omni 1.1 in continuous camera pushes connecting exterior and interior frames.
- Long, unbroken zoom-outs and continuous camera pulls with natural scale and zero jump cuts.
Availability, infrastructure and safety
- General availability: Gemini Omni 1.2 Flash is rolling out across Google’s developer and enterprise platforms via the Gemini API and Gemini for business.
- Infrastructure: the service is backed by Google Cloud’s enterprise-grade infrastructure and security. Provisioned Throughput is available.
- Outputs include an imperceptible SynthID watermark.
How to get started
- Developers using the Gemini API: prototype in Google AI Studio, explore interactive starter apps, and consult developer documentation and pricing.
- Businesses: try it in the Google Cloud console and build enterprise solutions with the API.
- Google product users: direct scenes and multi-turn edits in Flow; generate 15-second clips in Google Vids at no additional cost with a Google account or Google Workspace subscription; create 15-second videos in the Gemini app (rolling out to Google AI Ultra subscribers); and scale video campaigns in Google Ads.
Ongoing development areas
Google says work will continue to improve performance on fast sports action, rapid script changes, and complex physics, but positions Omni 1.2 Flash as suitable now for businesses, developers and creators producing video at scale.
Key facts
- Model: Gemini Omni 1.2 Flash
- Availability: generally available across Gemini API and Gemini for business
- Price: $0.07 per second for 720p (30% reduction vs Omni 1.1)
- Native clip length: up to 15 seconds (3–15 s selectable in 1 s steps)
- Short motion reference support: up to 3-second clips
- Output watermark: imperceptible SynthID
- Infrastructure: Google Cloud-backed, Provisioned Throughput available
All examples and capabilities described above are taken from Google’s announcement and the product documentation shared at launch.
