Google Cuts Gemini Omni 720p Video Price 30%, Extends Clips to 15 Seconds

A 15-second video generated at 720p with Google's newest model now costs $1.05. That figure comes from the pricing Google has attached to Gemini Omni 1.2 Flash, which the company has made generally available through the Gemini API and Gemini for business. The rate is $0.07 per second, which Google says is 30% lower than the price of Gemini Omni 1.1.

The release also lengthens what the model can make in one go. Native video generation grows from 10 seconds to 15, and users can set a duration anywhere from three to 15 seconds in one-second steps.

The larger change is in how clips get revised. Google describes a workflow for drafting, editing and upscaling, in which a user refines an existing video through follow-up instructions instead of starting over. The company says conversational editing works across full 15-second clips: swap a product or a background, change the lighting or weather, or alter the aspect ratio without regenerating the whole video. Teams can draft at 360p or 720p, adjust selected details, and then upscale chosen outputs to 1080p or 4K. Omar Pera, CPO at Magnific, framed the need this way: the first generation is a starting point, and professional teams need precision to turn it into a finished shot.

Generation can draw on character, product and style images, audio cues, and motion references of up to three seconds. Google reports better consistency in faces, clothing, packaging text and brand colors as the camera angle shifts. Its demonstrations include a virtual try-on sequence built from a model photograph and front-and-back images of three outfits, meant to show garment detail and the model's identity holding through turns and transitions. Another compares Omni 1.1 Flash with 1.2 Flash on a studio lighting poster, with the newer model assembling the poster's letters, materials and lighting details in sequence. A third is a four-shot, 15-second science explainer generated from a single visual reference, with explanatory labels built into the video.

Google says the update reduces visual drift, where objects change shape or appearance during motion and scene changes. That covers sequences generated between a supplied first and last frame, as well as extensions of existing clips. Dialogue has been revised too, with what Google calls more natural speech, clearer lip-sync and better assignment of distinct voices to different characters. One demo pairs a scene image with character, mouth and voice references to produce a 15-second animated conversation; another uses architectural photographs to guide a continuous camera move through the Pantheon.

Partners are building on this. Itay Schiff, Creative Director of Figma Weave, says teams there can direct an entire video sequence from a single generation and then refine it with references and branching versions on the canvas. Adobe plans to add Gemini Omni Flash 1.2 to Firefly, where creators already have access to Gemini Omni Flash, citing longer outputs and closer adherence to reference images.

Access differs by product. Developers can prototype in Google AI Studio and reach the model through the Gemini API, while businesses can use the Google Cloud console and build through the API. In Google Vids, 15-second clips are offered at no additional cost with a Google account or a Google Workspace subscription. Flow supports scene direction and conversational edits, and Google Ads offers a route to video campaigns. Provisioned Throughput is available through Google Cloud for organizations running the model at scale, and rollout to Google AI Ultra subscribers in the Gemini app, for 15-second videos, is under way. Outputs carry an imperceptible SynthID watermark.

Google acknowledges the limits: fast sports action, rapid script changes and complex physics are still in development. Those are the areas to watch, along with Adobe's Firefly integration, which has been planned but not yet delivered.