Google Gemini Omni 1.1 Flash is a fast, conversational video generation and editing model. Create from text, images, or video, then refine through natural-language prompts.
Everything Gemini Omni Flash can do
Built for cadence
Google's fast, low-cost video tier, capped at 10 seconds — the one to reach for when a dozen takes beat one careful one.
Open on one frame, land on another
Attach a start frame, add an end frame, and the motion between the two is filled in — 3 to 10 seconds of it, in the aspect you pick.
Put a face, a product or a place on screen
Up to six reference stills per shot, carried through the take so the thing you attached is the thing that shows up.
Rework footage you already have
Attach the source clip and describe the change: a new grade, a fix, a different look. No frames to key, no mask to draw.
Frequently Asked Questions
How long can a clip be?
3 to 10 seconds. Ten is the hard ceiling on every mode, and a prompt describing something longer is clamped to it rather than running over.
Does it generate sound?
No. Every Omni Flash route returns picture only. For dialogue, foley and score in the same pass, use Veo 3.1.
How many images can I attach?
Up to six on the Omni tab, beside the prompt. The Frames tab takes two — a start frame, and an optional end frame it moves towards.
What does the Video tab do?
It edits a clip you attach: the prompt describes the change to apply — a restyle, a fix, a reshape. That route carries no length or aspect of its own, so those chips step aside.
Which aspect ratios are supported?
16:9 or 9:16, on the Omni and Frames tabs.
How long can the prompt be?
Up to 10,000 characters on every route, so a full art-direction brief fits without trimming.
