Author: Nadali. Prepared with AI assistance.
On November 5, Google announced the rollout of Gemini Omni 1.2 Flash, including 15-second clips in Vids. The company also claims better retention of reference details. Nadali has not independently tested that quality.
Who benefits: a small team already making work videos in Vids. A longer clip can be tried for one complete scene, but does not replace checking what the viewer will see and understand.
What you can adjust
The generation guide describes Create mode: the Omni menu sets a duration of 3–15 seconds. Insert the finished clip into the project to retain access after the session.
That range applies to Create. Do not automatically apply it to Animate or Edit, whose help sections describe different limits. A new model also does not establish that the separate paused rollout of image-based avatars has resumed.
Check the account and quota first
In a separate November 5 announcement, Google said holders of a Google account or Workspace subscription can generate up to three 15-second videos per month at no cost. This is the announced offer, not a check of access in your account.
The availability table lists 200 AI video clip seconds per month for Business Starter and 500 for Business Standard and Plus. Administrators can disable Vids or generative features.
However, the generation page still gives a general allowance of 50 videos per month for most users. The table also lists 60 seconds per month for personal Google accounts. Google does not explain here how these quotas relate to the three no-cost videos. Do not assume the allowances stack or form one limit for every plan. Check the remaining allowance in your own account before a series of attempts; do not plan volume around a single help-page figure.
Google’s language guide does not separately list Ukrainian for video clip generation; it specifies English for other, unlisted Workspace Gemini features. Do not promise the team support for Ukrainian prompts or accurate Ukrainian speech without checking.
One reviewed video before a batch of scenes
Our proposed first-trial process:
- Define one action or idea the scene should communicate. Prepare a short approved script.
- Use your own or permitted materials without customer data. For a real product demonstration, compare shape, colour and included items with the original.
- Watch the entire clip with sound, including the final seconds. Check text, objects and the sequence of actions.
- Label synthetic material where viewers could mistake it for real footage. Keep instructions available as text too.
Consider a fictional café making an atmospheric opening for an internal video. A generated scene can serve as an illustration, but should not be presented as footage of the café’s actual kitchen. For training on specific equipment, use a verified real demonstration rather than movements invented by the model.
This is a proposed workflow, not a Vids test or a measured time saving. The takeaway: the new duration gives a scene more room; get one result accepted by a person before scaling up. For the text track, see our caption review guide.
Sources read on November 7, 2026. We did not test account-level access or generation. The cover is an original AI illustration, not the Vids interface or an Omni output.
