Google Launches Gemini Omni 1.2 Flash with Extended 15-Second Video and 30% Lower Cost

Image: Google · Source

Google's Gemini Omni 1.2 Flash enables native 15-second video generation with improved consistency, multi-turn conversational editing, and 30% cheaper pricing, facilitating scalable high-quality video production.

Google has introduced Gemini Omni 1.2 Flash, an updated AI video generation model available via the Gemini API, Google Vids, and Gemini for business. This new version extends native video clip length to 15 seconds, up from 10 seconds in the previous release, providing users greater flexibility to create complete narratives, product explainers, and social advertisements in a single generation.

A key advancement is multi-turn conversational editing, which allows users to refine videos interactively without restarting the process. For example, one can swap backgrounds, change lighting conditions, or adjust aspect ratios through sequential prompts, maintaining continuity in character movements and scene elements. This interactive editing is supported across full 15-second clips, streamlining variant production for diverse audiences and platforms.

Gemini Omni 1.2 Flash also offers enhanced reference adherence, maintaining consistency of characters, products, and branding elements such as faces, garments, typography, and colors across camera angles throughout the clip. This reliability supports use cases like virtual try-on sequences and branded video advertisements, where visual fidelity to source images is critical.

Visually, the model produces more stable and cinematic footage by reducing visual drift and jump cuts during complex camera movements like flythroughs and zoom-outs. Scene consistency is preserved frame-to-frame, even in extended shots involving multiple transitions and motions.

In addition, the model delivers more natural-sounding speech with improved lip-sync and multi-speaker dialogue capabilities, enhancing the realism of avatars and narrated presentations.

Google has also reduced the usage cost by 30% compared to Gemini Omni 1.1, pricing video generation at $0.07 per second for 720p resolution. Workflows emphasize cost-efficiency by enabling fast drafts at lower resolutions (360p or 720p), targeted refinements via multi-turn edits, and the option to upscale final selections to 1080p or 4K.

Gemini Omni 1.2 Flash is generally available now, backed by Google Cloud’s enterprise-grade infrastructure, with features like provisioned throughput and imperceptible SynthID watermarks on outputs. Users can access this technology through developer tools such as Google AI Studio, enterprise platforms, or consumer-facing apps like Google Vids and the Gemini app (rolling out to Google AI Ultra subscribers).

This release represents a significant step in AI-driven video content creation, supporting creators, developers, and businesses aiming to produce scalable, high-quality video content efficiently.