Google cuts Omni video pricing 30% and extends clips to 15 seconds Google cuts Omni video pricing 30% and extends clips to 15 seconds

A longer clip changes the unit of work

The new release lets callers choose duration in one-second increments. Google also offers follow-up instructions for changing an existing clip, including its lighting, background or product. That separates the initial generation from later revisions. A fifteen-second spot can now fit inside one ordinary generation, while the previous model stopped at ten seconds. This reduces the need to join shorter clips for that particular length. It does not turn a series of unrelated prompts into a continuous production. The practical change is a larger single-shot allowance alongside an editing conversation that can retain the earlier clip as its starting point.

Cloud access has a specific interface

The model documentation identifies gemini-omni-1.2-flash as a generally available release. It supports the Interactions API, with horizontal 16:9 and vertical 9:16 output. A prompt can include up to ten reference images, while an input video for editing can run up to fifteen seconds. These are separate controls: reference count describes how much visual guidance enters a request, and duration describes how long the resulting shot can run. The published model table also separates ordinary generation from first-and-last-frame generation and extensions. That distinction matters for applications which already assemble videos around fixed endpoints. A longer headline duration does not automatically lengthen every stage of such a pipeline.

Resolution and capacity remain separate choices

Google Cloud describes a workflow that starts with 360p or 720p drafts and upscales selected results to 1080p or 4K. This makes resolution another production decision after the initial composition. The service offers both usage-based access and Provisioned Throughput, giving applications distinct ways to obtain capacity. Google also applies SynthID watermarking to generated output. Its launch post includes endorsements from creative platforms already using the model. Those examples describe commercial integrations, rather than a controlled comparison of the quality or expense of finished videos.

Vids supplies a smaller entry point

Google is putting the release inside Vids as well as developer services. Its November 5 announcement gives Google-account holders and Workspace subscribers up to three fifteen-second generations per month at no cost. This allocation belongs to the video-creation application; it is not an API credit. For someone making an occasional presentation clip, that provides a different entry point from building a generation service. The distinction also changes the relevant purchase decision. A small monthly allowance can cover an experiment inside an editor, while recurring software-generated video depends on the paid model interface and its published limits.