Back to Models Guide
Googlevideo

Complete Guide to Using Gemini Omni Video

Try This ModelTutorial

Overview

Google Gemini Omni — natural-language video generation with native audio. Optional reference images, source clip, voices, and characters. Tier prices by resolution × duration.

Input Parameters

Prompt
textarearequired

Text prompt (max 20000 chars).

Reference Images (optional)
file

0–7 reference images for scene, style, or storyboard guidance.

Reference Clip (optional)
file

Optional source clip ≤30s, ≤100MB. Adding a clip causes the model to ignore the Duration setting.

Voices (optional)
multi-select

Up to 3 Gemini voices for narration or dialogue. Leave empty for silent video.

Options
AchernarAchirdAlgenibAlgiebaAlnilamAoedeAutonoeCallirrhoeCharonDespinaEnceladusErinomeFenrirGacruxIapetusKoreLaomedeiaLedaOrusPuckPulcherrimaRasalgethiSadachbiaSadaltagerSchedarSulafatUmbrielVindemiatrixZephyrZubenelgenubi
Characters (optional)
multi-select

Up to 3 of your saved Gemini Omni Characters. Or @mention them in the prompt.

Duration
selectrequired
Options
4s6s8s10s
Default: 8
Resolution
select
Options
720P1080P4K
Default: 720p
Aspect Ratio
select
Options
16:99:16
Default: 16:9
Seed
number
Min: 0Max: 2147483647Default: 0