Prompt to video
Define the subject, action, shot size, camera movement, setting, lighting, and ending state. Without video input, choose a duration of 4, 6, 8, or 10 seconds.
Z-Image
Gemini Omni
Multimodal video generation with references
Create from text, images, or references. Video input makes duration automatic; without video, choose 4, 6, 8, or 10 seconds.
0 chars
720p
1080p
4s
▭
16:9
▭
9:16
🎬
Generation Result
Create stunning videos from text or images instantly with Z-Video
Estimated time: 30-60s
👑Membership required for video generation
What Gemini Omni does on Z-Image
Gemini Omni is a multimodal video model that can begin with a prompt, a still image, several visual references, or an existing video clip. It is useful when motion and scene direction need to follow material you already have instead of starting from text alone.
This page currently exposes the Gemini Omni Video workflow at 720p or 1080p in 16:9 or 9:16. KIE also documents separate Audio and Character asset APIs, but those create reusable IDs for other workflows; they are not additional video models and are not enabled here. 4K is also deferred.
Choose the right video input
The input changes how much control you give the model. Decide whether you need an original scene, appearance continuity, or motion continuity before uploading references.
Image-led video
Use a still image to anchor a character, product, environment, or visual style. Describe motion and camera behavior rather than repeating only what is already visible in the image.
Video-guided creation
Upload one source video when its movement or scene progression should guide the result. The model determines output duration automatically, so the manual duration control is ignored in this mode.
Best Scenarios
Reference-led visual concepts
Short social clips
Capabilities
Text, image, and reference modes
Optional 1 video input; video uses 2 quota units and each image uses 1, with 7 total
Automatic duration with video; otherwise 4, 6, 8, or 10 seconds
Audio ID, Character ID, and 4K remain deferred
Recommended Setup
Use references to establish the subject and visual direction
Avoid video input when a fixed duration is required
Prompt Examples
A cinematic drone shot over a misty forest at sunrise, slow camera movement
Prompting tips for more coherent motion
Write the action as a short sequence with a clear beginning, change, and ending instead of listing unrelated visual ideas.
Name the camera behavior explicitly, such as locked shot, slow push-in, tracking shot, handheld movement, or overhead view.
Give every reference one purpose. A small, consistent reference set is usually clearer than filling all available slots with competing cues.
When using source video, describe what should be preserved and what should be transformed. Do not rely on a duration choice because video input makes duration automatic.
Gemini Omni questions
What is Gemini Omni best used for?
It is well suited to short campaign concepts, product motion, reference-led storytelling, scene transformations, social clips, and previsualization where text, images, or a source video need to guide one coherent result.
What input modes does this page support?
You can generate from text, animate a single image, or use multimodal references that may include images and one video. The controls shown on the page reflect the combinations currently supported by Z-Image.
Why does duration disappear when I upload a video?
With no video input, you can choose 4, 6, 8, or 10 seconds. When one source video is supplied, Gemini Omni determines the output duration automatically and the manual duration parameter no longer applies.
How are Gemini Omni reference limits calculated?
A request has seven reference units. Each image uses one unit and the single allowed video uses two units. The current interface enforces the remaining capacity as you add media.
Which output settings are currently available?
Z-Image currently supports 720p and 1080p output with 16:9 or 9:16 framing. Although upstream material mentions 4K, it is not enabled in this release.
How are Gemini Omni Audio and Character related to this video model?
They are separate asset-creation APIs that produce Audio IDs or Character IDs for compatible workflows. They are not standalone video generators, and Z-Image does not expose those ID-based features on this page yet.
Related models
Z-Image is an independent generation platform and is not affiliated with or endorsed by the model provider.
Popular Tools
Explore our most popular creative tools
Z-Image Edit
Upload image, transform with one sentence
Try it now
Creative Engine
One sentence, AI provides infinite prompt creativity.
Try it now
Image Parse
Upload image, retrieve prompt instantly.
Try it now
Prompt Library
Discover thousands of high-quality AI prompts.
Try it now
Z-Image LoRA
Combine multiple LoRA models to create unique AI artwork
Try it now
AI Image Generator
Turn your text into stunning images instantly.
Try it now
Style Library
Explore curated artistic styles for your creations.
Try it now
Remove Background
Instantly remove backgrounds from images with precision.
Try it now
Image Upscaler
Enhance image resolution up to 4K/8K.
Try it now
Image Reframe
Expand images to any aspect ratio with outpainting.
Try it now
Camera Angle
Create a new view with continuous 3D camera control.
Try it now
Image Layers
Split an image into editable transparent layers.
Try it now
Image Parallax
Turn one image into interactive parallax animation.
Try it now
Image to ASCII
Convert images into customizable ASCII art locally in your browser.
Try it now
GIF to ASCII
Turn animated GIFs into downloadable ASCII animations.
Try it now
Video to ASCII
Convert MP4, WebM or MOV clips into downloadable ASCII videos.
Try it now
Depth Map Video
Create a private depth map video locally in your browser.
Try it now