AI Pulse · Frontier Labs Watch

Google has updated its Gemini Omni video model with features that matter most to teams producing real creative work at volume. Gemini Omni 1.1 Flash can now extend a scene by up to 40 seconds, use keyframes to control transitions between shots, create low-cost 360p drafts for faster iteration, and upscale finished work to 4K. Google says the draft mode costs about one-third as much, which gives creative teams room to test more versions before committing to a final render. The model is being packaged into the Gemini Enterprise Agent Platform, not positioned as a casual consumer video tool. Read Google’s announcement.

For a business owner, the practical shift is not “AI makes a finished advert at the click of a button.” It is that video production teams can spend less time waiting on early drafts and more time deciding what message, sequence, and visual direction actually works. Keyframe-controlled transitions and scene extension are especially useful when a team needs continuity across a campaign, product demo, training video, or social cutdown.

The important operating question is where this fits in your approval process. Cheap drafts are useful only if someone can quickly review them against brand rules, campaign goals, and audience needs. Final 4K output still needs the same scrutiny as any customer-facing asset. This is the kind of thing we build into an AI command centre, where teams can track the tools, workflows, and controls that move from experimentation into daily production.