What happened

On 5 November 2026 Google announced Gemini Omni 1.2 Flash, the newest version of its Omni video generation and editing model. The announcement post, written by Google product manager Anish Nangia, says the model now makes native clips of up to 15 seconds (Omni 1.1 stopped at 10), follows reference images and clips more closely, and supports multi-turn editing, where a user changes a clip by describing the change in conversation. Google also says the price is 30% lower than Omni 1.1.

We read Google’s post on 6 November 2026. All figures below come from that post unless stated otherwise.

Key details

  • Length: clips run from 3 to 15 seconds, set in one-second steps. A reference motion clip can be up to three seconds long.
  • Price: Google lists $0.07 per second for 720p video, a 30% drop from Omni 1.1. By our own arithmetic, a full 15-second clip at 720p would cost about $1.05 at that rate; Google does not give a per-clip price or the exact Omni 1.1 rate.
  • Resolution: users can draft at 360p or 720p and upscale to 1080p or 4K.
  • Availability: Google says the model is generally available on the Gemini API (Google AI Studio) and in Gemini for business and Google Cloud, where Provisioned Throughput is offered. It is also in Flow, in Google Vids (15-second clips “at no additional cost” with a Google account or Workspace subscription), in Google Ads, and rolling out in the Gemini app to Google AI Ultra subscribers.
  • Provenance: Google says outputs carry a SynthID watermark.
  • Not stated: the post gives no benchmark scores (its Omni 1.1 comparisons are side-by-side example clips) and no license or usage terms.

Why it matters

For anyone who pays for AI video by the second, this is a price cut and a longer clip on a model that Google now says is generally available, not in preview. At $0.07 per second a 15-second shot costs roughly a dollar at 720p, and the draft-then-upscale option means experiments can stay at the cheaper setting. Putting 15-second clips into Google Vids without an extra charge brings the model to people who never touch an API.

The caveat is that quality claims rest on Google’s own examples. Without published benchmarks or independent tests, it is hard to say how Omni 1.2 Flash compares with other video models at the same price. Check the per-second rate for the resolution you actually need before building on it.

Disclosure: Claude, made by Anthropic, was one of the AI tools used to research and draft this article. Google is not affiliated with Deyron Labs; all product claims are Google’s own.

Sources

  1. Introducing Gemini Omni 1.2 Flash Primary source , Google
  2. Google DeepMind blog listing (Gemini Omni 1.2 Flash) Google DeepMind