Google DeepMind announced on Nov. 5, 2026, the rollout of Gemini Omni 1.2 Flash, an updated version of its video generation model that introduces native 15-second video creation, improved reference adherence, and multi-turn conversational editing. The company is also lowering the model's price by 30% compared with Gemini Omni 1.1, bringing the cost to $0.07 per second for 720p video.

The model is generally available starting today on the Gemini API and Google Cloud for production needs, as well as through Gemini for business. According to the announcement from Anish Nangia, Product Manager at Google DeepMind, the update lowers costs while providing better consistency for professional projects, and the capabilities can be used across the Gemini API, Google Vids, and Gemini for business to produce high-quality video content at scale.

The headline changes in the release are:

• Native support for 15-second video creation, up from 10 seconds in Gemini Omni 1.1
• Multi-turn conversational editing that allows faster adjustments without starting over
• Improved reference adherence for characters, products, and styles
• A 30% price reduction to $0.07 per second for 720p video
• General availability on the Gemini API and Google Cloud

Fast and cost-effective at scale

Beyond the reduced per-second price, Google said the surrounding workflow is structured to keep costs down. Users can draft clips quickly at 360p or 720p resolution, fix individual details with a follow-up edit rather than re-generating an entire clip, and then upscale their best creations to 1080p or 4K. The company framed the combination of lower pricing and an edit-first workflow as a way to produce high-quality video content at a lower overall cost. For teams budgeting output, the numbers translate directly: the 30% cut implies Gemini Omni 1.1 was priced at $0.10 per second, meaning a max-length 10-second clip cost $1.00 under 1.1, while 1.2's 15-second ceiling now costs $1.05 — 50% more runtime for 5% more spend at the cap. Because billing follows output length, refining existing clips through follow-up edits rather than re-generating them is the other lever Google pointed to for keeping total production costs down.

Longer clips up to 15 seconds

Gemini Omni 1.2 Flash can generate and edit videos natively up to 15 seconds long, an increase from the 10-second limit of Omni 1.1. Duration can be specified in one-second increments from 3 to 15 seconds, which Google said gives users room to tell complete stories, create product explainers, or produce full social ad spots in a single generation.

Better reference adherence

Users can guide generation by combining character, product, or style images, short motion clips of up to three seconds, and audio cues. According to Google, Omni 1.2 Flash holds onto those details from start to finish: faces, garments, packaging typography, and brand colors stay consistent across camera angles. The company said this makes it easier to turn existing product photography or character sheets into virtual try-on clips, catalog spins, and branded video ads.

In one example described in the post, combining a single model photo with front-and-back product shots of three outfits allows the model to generate a 15-second virtual try-on sequence across 360-degree turns and match cuts while keeping the model's identity and each garment's cut, drape, and back details consistent.

More consistent results

Google said the model maintains stronger character, object, and scene consistency from the first frame to the last, with less visual drift or morphing across motion and camera transitions. Whether interpolating between a first and last keyframe or extending a clip to build a longer sequence, products, characters, props, and environments stay consistent from shot to shot.

The company illustrated the difference with a comparison built on the same static studio-lit exhibition poster and a five-layer animation prompt. According to the post, Gemini Omni 1.1 Flash does not match the reference image as closely as version 1.2 does. Gemini Omni 1.2 Flash builds the poster sequentially, layer by layer, on the dusty-pink cyclorama, raising the hexagonal plinth and assembling the ribbed glass, warm filament, faceted crystal, and coiled-cord letterforms spelling HALO, GLOW, and LUMEN, while locking every material, shadow, and typographic detail consistently through the final frame.

Easier editing in conversation

Users can refine the same clip turn by turn — swapping a product or background, changing the lighting or weather, or switching the aspect ratio — instead of starting over from scratch. Multi-turn conversational editing works across full 15-second clips, which Google said makes it easy to produce ad variants for different audiences and placements without re-briefing or re-generating.

In a demonstration cited in the announcement, starting from one generated clip, a sequence of conversational prompts swaps the background turn by turn — from a striped pavilion tent to an oak forest, a stable with a white horse, a market fair, and a stone castle — while keeping both characters' performances, armor reflections, and camera framing intact.

More natural speech

The model generates scenes with more fluid, natural-sounding dialogue, clearer lip-sync, and distinct multi-speaker conditioning, according to Google. The company said these improvements show up whether users are directing multi-character dialogue in Flow, creating video explainers in Gemini for business, or producing polished, professional videos with avatars, music, and narration in Google Vids.

In one example, conditioned on a single 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.

More controllable, cinematic look

Google said users can direct complex, continuous camera movements — from sweeping FPV flythroughs to dramatic zoom-outs and orbits — with significantly fewer unwanted jump cuts, steadier scale, and a more polished, cinematic look.

In a showcase example, guided by two architectural reference photos, the model directs one continuous FPV drone flight across the piazza, through the bronze doors, and up through the Pantheon's coffered dome to the oculus with zero jump cuts.

Longer clips with smoother motion and tighter consistency

Gemini Omni 1.2 Flash generates clips up to 15 seconds long that hold references together better than before. Google said characters, text, and camera moves stay consistent across longer shots, multiple turns, and transitions with no cuts, and highlighted several demonstrations:

• 15-second long clips: A single style frame is turned into a four-shot, 15-second macro science explainer with accurately spelled, motion-tracked callouts.
• Multi-turn motion transfer: The model generates a rippling octopus in the first turn, then replaces it in the second turn with a dancer whose gown mirrors the exact fluid motion.
• First-to-last frame consistency (1.1 vs. 1.2): The model connects an exterior start frame and an interior end frame in one continuous camera push in 1.2 Flash. Google noted that subtle differences remain visible, such as the cats on the roof and a soup ladle turning into chopsticks.
• Unbroken crash zoom-out: The camera pulls continuously from a tight shoreline close-up to an aerial island vista in a single 10-second shot with natural scale and zero cuts.

Real-world use and enterprise availability

According to Google, businesses, startups, and creatives are already using Gemini Omni 1.2 Flash in their workflows. For the first time, Gemini Omni is now generally available, backed by Google Cloud's enterprise-grade infrastructure and security. Provisioned Throughput, Google Cloud's option for securing committed model capacity for sustained workloads, is available, and outputs carry an imperceptible SynthID watermark, Google DeepMind's technology for marking AI-generated content.

Getting started

Google said development is ongoing to further improve areas including fast sports action, rapid script changes, and complex physics — its own accounting of what still needs work, and a concrete reference point for teams weighing the model against demanding production scenarios. The company positioned Gemini Omni 1.2 Flash as well suited for businesses, developers, and creators producing videos at scale, with availability rolling out across Google's products and developer platforms:

• Developers using the Gemini API can prototype in Google AI Studio, explore interactive starter apps, and consult the developer documentation and pricing.
• Businesses can try the model 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, generate 15-second clips at no additional cost in Google Vids 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.

The full announcement is available on the Google DeepMind blog: Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips.