Google has released Gemini Omni 1.2 Flash, a video generation model now available via the Gemini API and Gemini for Business. The update lowers the cost to $0.07 per second for 720p video and extends native generation to 15 seconds.
Developers can now draft at 360p or 720p, correct specific details without re-generating the whole clip, and upscale the final result to 1080p or 4K.
Extended clip duration
Generation and editing now support clips up to 15 seconds, an increase from the 10-second limit in the previous version. Users may specify durations in one-second increments between three and 15 seconds. This allows for complete narratives, product explanations, or full social media advertisements to be produced in a single pass.
Improved reference adherence
Generation responds more accurately to provided images, short motion clips of up to three seconds, and audio cues. Character faces, clothing, packaging typography, and brand colours remain consistent across different camera angles. This capability simplifies the creation of virtual try-on clips, rotating product shots, and branded video ads from existing photography.
When provided with one model photo and front-and-back shots of three outfits, the model produces a 15-second virtual try-on sequence covering 360-degree turns and match cuts while maintaining the model’s identity and the cut, drape, and back details of each garment.
Reduced visual drift
Character, object, and scene consistency improves from the first frame to the last, reducing visual morphing during motion and camera transitions. Whether interpolating between keyframes or extending a clip, products, characters, props, and environments stay consistent shot to shot.
Using a static studio lighting exhibition poster and a five-layer animation prompt, Gemini Omni 1.2 Flash assembles the poster layer by layer on a dusty-pink cyclorama. It places the hexagonal plinth, ribbed glass, warm filament, faceted crystal, and coiled-cord letterforms for HALO, GLOW, and LUMEN, locking every material, shadow, and typographic detail through the final frame.
Iterative conversational editing
Users can refine the same clip turn by turn. Swapping a product or background, changing lighting or weather, or switching the aspect ratio no longer requires starting over. Multi-turn editing works across full 15-second clips, allowing for ad variants tailored to different audiences and placements without re-briefing.
Starting from one generated clip, multi-turn prompts swap the background sequentially from a striped pavilion tent to an oak forest, a stable with a white horse, a market fair, and a stone castle. Both characters’ performances, armor reflections, and camera framing remain intact.
Natural speech and lip-sync
Generated scenes feature fluid dialogue, clearer lip-sync, and distinct multi-speaker conditioning. This improvement applies to directing multi-character dialogue in Flow, creating video explainers in Gemini for Business, or producing polished videos with avatars, music, and narration in Google Vids.
Conditioned on one conservatory scene image alongside four character, mouth, and one-second voice references, Gemini Omni 1.2 Flash routes each sampled voice to the correct character with expressive lip-sync and natural comedic timing across a 15-second animated scene.
Controlled cinematic movement
Complex, continuous camera movements such as sweeping FPV flythroughs, dramatic zoom-outs, and orbits produce fewer unwanted jump cuts. Scale remains steadier, resulting in a more polished look.
Guided by two architectural reference photos, Gemini Omni 1.2 Flash directs one continuous FPV drone flight across a piazza, through bronze doors, and up through the Pantheon’s coffered dome to the oculus with zero jump cuts.
Consistency in longer shots
Clips up to 15 seconds hold references together better than before. Characters, text, and camera moves stay consistent across longer shots, multiple turns, and transitions without cuts.
Availability and rollout
Businesses, startups, and creatives are already using Gemini Omni 1.2 Flash in their workflows. For the first time, Gemini Omni is generally available, backed by Google Cloud’s enterprise-grade infrastructure and security. Provisioned Throughput is available, and outputs carry an imperceptible SynthID watermark.
Development continues to improve fast sports action, rapid script changes, and complex physics. The current version is ideal for businesses, developers, and creators producing videos at scale.
Access is available through the following channels:
- Developers using the Gemini API can prototype in Google AI Studio, explore interactive starter apps, and check the developer documentation and pricing.
- Businesses can try it in the Google Cloud console and build enterprise solutions with the API.
- Users of Google apps can direct scenes and multi-turn edits in Flow. Google Vids generates 15-second clips at no additional cost with a Google account or Google Workspace subscription. The Gemini app creates 15-second videos for Google AI Ultra subscribers. Video campaigns can be scaled in Google Ads.
