Google multimodal video model
Gemini Omni Video Generator
Animate a reference image with Gemini Omni and describe the motion, camera, sound, and scene changes you want in one prompt.
Available on FreeWanVideo: FreeWanVideo exposes Gemini Omni as an image-to-video workflow with 4-, 6-, 8-, and 10-second choices, landscape or portrait framing, and 720p or 1080p controls.
Current site capabilities
What you can create with Gemini Omni
Multimodal generation
Gemini Omni can reason across text, images, audio, and video for cohesive short-form generation.
Image-led motion
Upload a photograph, illustration, or product image and describe how it should come to life.
Short-form duration control
Choose 4, 6, 8, or 10 seconds in the current FreeWanVideo interface.
Verified model facts
Official Gemini Omni specifications
Official family specifications are shown separately from the options currently exposed by FreeWanVideo.
- Official model code
- gemini-omni-flash-preview
- Official output
- 3–10s at 720p
- Core workflow
- Generate and edit
Google released Gemini Omni Flash as a preview video generation and editing model in June 2026.
The Gemini API model card lists 3- to 10-second video output at 720p and 24 FPS.
Google documents text-to-video, image-to-video, reference-to-video, and conversational editing tasks.
Use cases
Best for
- Fast animation from a reference image
- Short landscape or portrait clips
- Prompts combining visual, motion, and sound direction
Direct answers
Gemini Omni FAQ
Is Gemini Omni an official Google video model?
Yes. Its official preview name is Gemini Omni Flash, with the model code gemini-omni-flash-preview. FreeWanVideo shortens the display name to Gemini Omni.
Can Gemini Omni animate an image?
Yes. Google's official documentation supports image-to-video, and the FreeWanVideo page opens directly in an image-led workflow.
What output does the official Gemini API document?
The official Gemini Omni Flash model card currently lists 3- to 10-second video output at 720p and 24 FPS. FreeWanVideo's provider interface may expose additional controls separately.
Explore more AI video models
Choose a model that matches your input, resolution, audio, and speed needs.
Official sources
Model facts reviewed July 26, 2026. Product availability on FreeWanVideo can differ from the full official model family.