Google's Gemini Omni 1.2 Flash: 15-Second Video Generation at 30% Lower Cost

On November 5, Google released Gemini Omni 1.2 Flash, a multimodal model that can generate video clips up to 15 seconds in a single request — an increase from the previous 10-second limit of Omni 1.1.

Pricing and Billing

At 720p resolution the price is $0.07 per second (7 ¢/sec), which is 30 % lower than the prior version. Billing is per second, with duration selectable in 1-second increments from 3 to 15 seconds. Drafts can be produced at 360p or 720p; upscaling to 1080p or 4K is possible after approval, though the announcement only specifies the 720p rate, so upscale costs should not be estimated at $0.07/sec.

Iterative Editing and Reference Controls

Unlike the previous version, Omni 1.2 Flash supports multi-turn conversational editing . A generated clip can be refined in successive turns — changing the product, background, lighting, weather, or aspect ratio — without re-generating the entire sequence. The model accepts:

Reference images for people, products, and style to preserve faces, clothing, packaging text, and brand colors across angle changes.

Motion reference clips up to 3 seconds to dictate movement trajectories.

Audio prompts that carry tone; lip-sync follows the generated audio, and multiple speakers can be handled separately.

Google states that dialogue flows more naturally, lip-sync is clearer, and unwanted jump cuts during continuous camera moves are reduced.

Watermarking and Enterprise Features

Every output carries an imperceptible SynthID watermark . Enterprises can provision reserved throughput for predictable capacity.

Current Limitations

Google acknowledges that fast-paced sports, rapidly changing dialogue, and complex physics are still being improved. The 15-second ceiling is a maximum; not every clip needs to use the full duration.

Availability and Integration

Developers access the model via Gemini API through AI Studio .

Enterprises use the Google Cloud console .

Google’s own products — Flow, Vids, the Gemini app, and Google Ads — have already integrated this version.

Free Tier in Google Vids

Users with a Google account or Workspace subscription can create 3 videos of 15 seconds each per month at no cost in Google Vids. In the Gemini app, the 15-second capability is rolling out to AI Ultra subscribers gradually. API usage is billed per second, while Vids uses a separate monthly quota; the two billing paths are independent.

Suggested Workflow for Product Videos and Ad Variants

Define the required duration based on the action script and generate the base clip.

In subsequent conversation turns, swap backgrounds, replace products, or adjust lighting on the already-generated footage.

Keep drafts at 720p to control costs; lock in the second count within budget before upscaling.

Sources

Gemini Omni 1.2 Flash announcement: https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemini-omni-1-2-flash/

Google Vids integration post: https://blog.google/products-and-platforms/products/workspace/omni-1-2-flash-google-vids/

Code example

👀点击关注
AI前沿快速Follow