Google Cloud announced the general availability of Gemini Omni 1.2 Flash, a video generation and editing model aimed at the high-volume business video use cases that companies produce every day: product videos, virtual try-ons, and ads. The model is available to developers through the API, to business users through Gemini for business, and inside Google Vids, which lets teams combine generated clips with avatars, music, and voiceover for explainers, teasers, social posts, and sales pitches.

Compared with the previous Flash version, the main changes are cost and capability. Video is now priced at $0.07 per second of 720p output, about 30% less than Gemini Omni 1.1 Flash, with similar savings at other resolutions. Generated and edited clips can be up to 15 seconds long, up from 10 seconds. The model has stronger adherence to image and video references, which is intended to keep products, characters, and brand details consistent from shot to shot. Users can draft at 360p or 720p and then upscale the best results to 1080p or 4K. Speech generation now supports more fluid dialogue, clearer lip-sync, and distinct multi-speaker conditioning. The same clip can also be refined conversationally, for example by swapping a product, changing lighting, or changing camera angle, instead of re-generating from scratch. Audio references are explicitly gated behind allowlisting.

For production use, the release is backed by Google Cloud‘s enterprise infrastructure and security, offers Provisioned Throughput, and all outputs carry an imperceptible SynthID watermark. The post frames this as a production-ready model, not a research preview.

Several creative platforms are integrating Gemini Omni 1.2 Flash directly into their products. Artlist cites reference-based generation, video extension, and creative consistency. FLORA says the model edits exactly what is asked, such as a colorway or background, while leaving the product untouched, and that its video-to-video capability can take footage into entirely new styles. invideo reports that internal evals show strong prompt adherence, realistic physics and sound, and that 15-second generation means a full scene in one pass. Krea highlights speed and accuracy, with outputs keeping detail and changing only what is requested, which the company plans to use for character and scene replacements. Picsart mentions that draft mode lets teams iterate concepts in minutes and then render in 4K for TV-ready output, with crisp text and consistent characters, now available to over 100 million users. Runway is also using the model for enterprise-scale professional video production.

Retail and marketplace use is a distinct theme. Photoroom chose Gemini Omni Flash to power its Video Ad Maker and Product Animator, citing the balance of quality, speed, and cost; version 1.2 adds 15-second videos, a top user request, and the general availability plus SLAs make it viable for production. QuickFrame AI uses the 15-second generation at 1080p to create a complete connected-TV ad in a single generation, which it says makes TV ad creation simpler for businesses of all sizes. Retailers and marketplaces are using the model to turn existing product photos into product spins, lifestyle clips, try-on videos, and video ads at catalog scale.

Gemini for business users will soon be able to create a short video from a prompt or a product photo directly in chat. Developers can try the model in the Google Cloud console, build with the API, and review pricing on the published page. The post closes with a request for feedback, noting that the model continues to be improved based on what builders report.