Gemini Omni Flash and the Gemini Omni models

Gemini Omni is Google's second video line after Veo: a conversational model that generates and edits video inside a chat rather than from a one-shot prompt, announced at I/O 2026. Two versions are on fal: the original Gemini Omni Flash and Gemini Omni Flash 1.1, which followed on the Gemini API in August 2026 with scene extension and reference clips and reached fal as separate endpoints with a 360p draft tier and 4K. Both pages run the same nine clips as every other model here.

Google DeepMind · facts checked

Which one to pick

Side by side, measured

ModelPriceSeconds to make one clipCost of its fixed setLength
Gemini Omni Flash 1.1$0.10 per second at 720p
via fal
31 to 60$4.50
9 clips
3 to 10 seconds in whole seconds; fal's default is 8
Gemini Omni Flash$0.10 per second of 720p video
via Gemini API
36 to 64$5.85
9 clips
3 to 10 seconds

Seconds and cost are from our own runs of each model's fixed set: the same 9 clips for every model that makes clips from a prompt. Open a model for every clip and the request behind it.

Every version

VersionReleasedWhat changed
Gemini Omni Flash
gemini-omni-flash
Announced at Google I/O 2026 and rolled to Gemini app users the same day; 720p clips of 3 to 10 seconds with audio, edited conversationally.
Gemini Omni 1.1 Flash
gemini-omni-1.1-flash
Generally available on the Gemini API: scene extension, first-and-last-frame interpolation, reference clips, selectable resolutions. On fal as its own v1.1 endpoints, $0.10 a second at 720p.

Frequently asked questions

Gemini Omni or Veo?

Veo 3.1 for a finished shot from one prompt, especially at 1080p or 4K. Gemini Omni when you want to talk a clip into shape over several turns. The two pages run the same nine clips.

Is Gemini Omni Flash free?

Not on the API, where Google lists no free tier. In the Gemini app it is part of the paid plans' video allowance.

What changed in Gemini Omni Flash 1.1?

On fal: a lower 720p price, a 360p draft tier, 1080p and 4K output and an optional last frame on image-to-video. On our nine prompts it spelled the title card and kept the pour intact, but it still gave our handleless dripper a handle in image-to-video, as version 1 did.