Gemini Omni Flash

Integrate text, images, and audio, and rapidly generate realistic videos and synchronized audio that comply with the laws of physics. It also supports advanced video editing using natural language and maintaining character consistency.

Text Friendly
Audio Support
Commercial Use

Input

Reference image*

Drag and drop media files from your computer, paste from clipboard (Ctrl/Cmd+V), or provide a URL. Accepted: .jpg, .jpeg, .png, .webp

Result

Idle

Ready to Generate

Configure your inputs and click run to generate an image preview.

Gemini Omni Flash Model Introduction

Model Overview

Gemini Omni Flash is Google’s advanced, next-generation high-speed multimodal video generation and editing model. It does not merely combine diverse assets such as text, images, audio, and existing videos, but also deeply infers the context of each, integrating them into a coherent and seamless video. Through advanced physical simulation, it reproduces gravity and fluid movements with extreme naturalness. Moreover, since it incorporates extensive knowledge about the real world, it can create realistic scenes without the need for detailed instructions. The model’s strength lies in its flexible video editing capabilities through natural language instructions. It allows users to repeatedly refine videos by adding or removing objects, changing camera angles, adjusting ambient lighting, and more, while maintaining consistency of characters. It can output videos with audio up to 10 seconds long, expanding creators’ expressive potential. From rapid verification of advertising creatives to short videos for social media and productions utilizing digital avatars, it strongly supports professional production workflows.

Pricing

ResolutionCredits Consumed
720p(credits/s)4

Technical Specifications

ParameterSpecification
Core CapabilityReference image to video
Resolution720p
Aspect Ratio16:9,9:16
Duration3,4,5,6,7,8,9,10

Explore similar models