Gemini Omni
A multimodal AI video generation model for text, image, and video source inputs.
Preview
More Video Models
Continue with other AI video generation models supported by AI Image Editor.
Gemini Omni
Generate short videos, image-to-video, and video editing drafts with Gemini Omni
Gemini Omni turns text ideas, reference images, or one source video clip into previewable AI videos. Describe the scene, action, camera, style, and pacing, then choose duration, aspect ratio, and resolution for the target channel.
Text to videoImage to videoVideo editing4K output
Prompt + image / video
Supports prompts, up to 7 reference images, and one source video clip.
AI video
Generated videos are saved to account history for preview, review, and download.
Gemini Omni video generation capabilities
The Gemini Omni page focuses on AI video generation, image-to-video, and source video editing for short video tasks that need natural-language control over visuals, motion, and camera direction.
Generate video from text
Enter the subject, action, scene, camera movement, lighting, style, and pacing so Gemini Omni can generate a short video concept from the prompt.
Animate image references
Upload product images, character images, scene images, or storyboard sketches so image-to-video results stay closer to existing visual assets.
Source video clip guidance
When existing motion, composition, or camera direction matters, upload one source video clip and describe the scene, style, and action you want to change.
Duration and aspect ratio controls
Create 4, 6, 8, or 10 second outputs and choose landscape or portrait for ad drafts, social clips, product demos, and content previews.
Resolution and credit estimate
Choose 720p, 1080p, or 4K; before submission, the page estimates credits based on resolution, duration, and whether video input is included.
Saved generation history
Completed videos are stored in generation history so you can revisit versions, check details, and download usable results.
What videos Gemini Omni is suited for
Use Gemini Omni when a video needs to combine prompt direction, reference images, or an existing video clip into a reviewable short video result.
Generate short videos from text ideas
Turn ad ideas, product concepts, story beats, or social scripts into AI video drafts with camera direction.
Write the camera language
Describe subject, action, environment, shot size, camera movement, lighting, visual style, and mood.
Match publishing specs
Choose duration, orientation, and resolution for the target placement to avoid heavy cropping later.
Check visual consistency
Preview characters, products, motion paths, text, brand elements, and rights risks before publishing.
Create image-to-video with references
Upload product, character, scene, poster, or storyboard images so Gemini Omni can reference the subject and visual style while generating video.
Provide visual anchors
Use images to define character appearance, product shape, scene atmosphere, color, and composition direction.
Explain what to preserve
Tell the model which details should remain recognizable in motion, such as product shape, character traits, or brand colors.
Compare versions quickly
Change the prompt, reference image set, aspect ratio, or resolution to get candidate videos with different styles and camera directions.
Recreate from a source video clip
Upload one short video clip as motion and composition reference, then use a prompt to change scene, material, atmosphere, or camera treatment.
Preserve original motion
When action timing, camera direction, or composition comes from existing footage, source video input is more direct.
Focus the specific change
Describe the background to replace, action to adjust, target style, lighting change, and elements that should remain.
Save and review output
After generation, review the video in history, compare versions, and download results suitable for editing or campaign use.
How to use Gemini Omni
01
Choose a video generation method
Start from text-to-video, image-to-video, multiple reference images, or a source video clip, depending on the material you have.
02
Write the prompt and set parameters
Describe the scene, subject, action, camera, style, and elements to preserve, then choose duration, orientation, and resolution.
03
Generate, preview, and download
Submit the task, wait for processing, preview the result in generation history, check visuals and details, then download the video.
Gemini Omni FAQ
What is Gemini Omni?
Gemini Omni is a multimodal video generation model in the Google Gemini family. In AI Image Editor, the Gemini Omni page supports text-to-video, image-to-video, reference image generation, and source video clip editing.
Which inputs does Gemini Omni support?
The current generator supports text prompts, image assets, and video assets. You can upload up to 7 reference images; when a source video is used, the system takes one video clip as input and guides video editing with the prompt.
Which durations, aspect ratios, and resolutions are supported?
The current page supports 4, 6, 8, and 10 seconds, landscape and portrait output, plus 720p, 1080p, and 4K resolution.
Where are results saved?
After generation, videos are saved to AI Image Editor generation history. You can preview, review, compare versions, and download results. Before publishing, check people, products, text, motion details, and rights risks.
Generate with Gemini Omni
Add a prompt and media to create short video drafts.
Mode Text / imageReferences Up to 7