Gemini Omni

Multimodal video generation — text, up to 7 reference images, or a video clip. Native 4K output and up to 10-second clips.

Max Duration: 10 seconds max
Resolution: 720p, 1080p, 4K
Aspect Ratios: 16:9, 9:16

4K max resolution10 sec max durationUp to 7 reference imagesT2V · I2V · V2V input modes

Create with Gemini Omni View Pricing

What makes Gemini Omni unique

Up to 7 reference images for character / scene continuity

Video-to-video editing with sub-clip trim range (up to 10s)

Native 4K output

Three input modes in one model: text, images, and video

Where it falls short

–Video-to-video output duration is decided by the model, not the prompt
–Aspect ratios limited to 16:9 and 9:16
–Image and video references share a 7-slot quota (each image = 1, each video = 2)

Best for

Reference-driven storytelling with consistent charactersShort cinematic clips at 4KRiffing on a source clip to expand a sceneMulti-reference compositions across characters and locations

Prompt tips

1.
Use reference images for identity. One image for the main character, others for wardrobe or location, instead of overloading a single image.
2.
When using a reference clip, trim to the moment you want the model to extend — short, focused clips give the cleanest continuation.
3.
If you need 4K, write longer prompts that describe lighting and texture; the model has more pixels to fill and richer prompts hold up better.
4.
Skip the duration field when using video-to-video — the model picks the length based on your clip.

How Gemini Omni works

Write a prompt describing the scene, camera, and style. Add up to 7 reference images for characters, products, or look-and-feel, or attach a short reference clip with start/end timestamps for video-to-video. Pick 720p, 1080p, or 4K and a duration of 4, 6, 8, or 10 seconds, then generate. Duration is fixed by the model when you provide a reference video.

How to use on Flashloop

Step 01

Select Gemini Omni

Open Flashloop and choose Gemini Omni from the model selector in the video creator.

Step 02

Write your prompt

Describe what you want to see in your video — be as detailed as you like.

Step 03

Generate & download

Hit generate and your video will be ready in seconds. Download or share directly.

Frequently asked questions

Three: pure text, up to 7 reference images for image-to-video, or a single reference clip with a trim range for video-to-video. You can mix images with a video reference as long as you stay under the 7-slot quota (each image is 1 slot, the video is 2).

Yes. Gemini Omni outputs natively at 720p, 1080p, or 4K. 4K costs more per generation since rendering is heavier on Google's side.

4, 6, 8, or 10 seconds. When you attach a reference video, the model picks the output duration itself and your duration choice is ignored.

16:9 landscape and 9:16 portrait. There's no square or other ratio at the moment.

Each reference image consumes 1 slot, and a reference video consumes 2 slots. The total must stay at or below 7 — so you can use 7 images alone, or 5 images plus 1 video, or 1 video by itself.