# Gemini Omni Video Canonical page: https://fuser.studio/models/gemini-omni-video Gemini Omni Video in Fuser generates 3-10 second video at up to 4K from text, images or source video. Direct shots and edit clips on one canvas. ## Specs | Spec | Value | | --- | --- | | Inputs | Prompt (text), Images (up to 5 images), Video | | Output | video | | Duration | 3–10 seconds | | Resolution | 360p, 720p, 1080p, 4K | | Cost | 58–8,290 credits per run | Grounded, sound-on clips and plain-language edits that keep your character and scene intact from the first take to the last Creator: Google DeepMind / Google [Roll the shot](https://app.fuser.studio) ## How Gemini Omni Video works You work like a director: write the shot, add the references or footage that anchor it, then set the length and frame. Draft cheap, and commit once the take holds. ### Write a shot plan with sound Name the subject, one or two actions, the camera move and the audio: quoted dialogue, ambience, effects, or 'no music'. ### Attach references or footage Add up to five reference images to lock a character or place, or drop in a clip of up to 10 seconds and describe only the change. ### Set length, frame and resolution Pick 3 to 10 seconds, 16:9 or 9:16, and draft at 360p before committing the winning take to a higher tier. ## What Gemini Omni Video is good at Two ways in: reference images that lock a character, product or place, and existing footage you revise by describing only the change. Either route comes back with synced audio built in. ### Sound baked into the shot Room tone, foley, music and lip-synced speech arrive with the picture. Put dialogue in quotation marks and name ambience in the prompt, or write 'no music' to keep it dry. ### Reference-locked characters and objects Feed up to five clean reference images and say what each one is for. The model holds the character, product or location steady while the camera moves and the action plays out. ### Edit footage by describing the change Hand it a clip of up to 10 seconds and write only the swap: a new background, a different season, a changed object. Lighting, reflections and water are re-simulated rather than pasted on, and the edit keeps the source duration. ### Cheap 360p drafts, finished renders Test composition, motion and character consistency at 360p, then re-run the winner. 720p is the native look; 1080p and 4K are upscales, and 9:16 is built in for vertical spots. ## Made with Gemini Omni Video Short, naturalistic clips with synced sound: industrial documentary, luxury product, weightless surrealism, vertical social motion and light-led architecture. ### Grounded physics and particle simulation in industrial documentary capture ### Material fidelity and tactile lighting in a luxury commercial ### Unbroken zero-gravity camera tracking across surreal landscapes ### High-speed vertical particle dynamics for social storytelling ### Consistent ambient light and spatial depth in architectural walkthroughs ## What people build with Gemini Omni Video Clips run 3 to 10 seconds, and a 360p draft tier makes testing cheap, so ad, product, previs and social teams can judge pacing first and spend on a higher resolution only for the winner. ### Ad teams re-dressing footage Take an approved 10-second spot and ask for a new background, a different season or a swapped product. The change-only prompt lets the original camera move and performance carry through. ### Product teams locking a hero object Load up to five clean references of a product and get consistent motion shots with native sound, from a latch click to a hinge swing, without rebuilding the object each time. ### Directors storyboarding at 360p Block out a talking scene with dialogue and ambience at draft cost, judge the pacing and character consistency, then commit only the best takes to a final render. ### Vertical spot makers with sound Build 9:16 clips for social feeds with the audio already in place. One clear action per beat, 3 to 10 seconds, and no separate sound pass. ### Interior walkthrough previews Preview how light moves through a space with a steady, naturalistic camera and room tone, a quick way to test mood before a full visualization pass. ## Gemini Omni Video questions ### How long can a Gemini Omni video be? Generated clips run from 3 to 10 seconds in whole-second steps. Source clips for editing are also limited to 10 seconds in public preview, and an edit keeps the source's duration, so the duration control is ignored in that mode. ### Does Gemini Omni generate sound with the video? Yes, audio is generated with the picture: ambience, effects, music and speech. Put spoken lines in quotation marks for lip-synced dialogue, name the sounds you want, or write 'no music' or 'no dialogue' to exclude them. If you plan to score a silent clip yourself, this is not the best fit. ### What resolution does Gemini Omni actually render at? 720p is the native generation. The 1080p and 4K options are upscaled from it, so they look sharper but not more detailed. A 360p draft tier is there for cheap iteration, and cost rises with resolution and duration. ### Can I edit an existing video with it? Yes. Upload a clip of up to 10 seconds and describe only the change, such as a new background or a swapped object. Swaps work best when the new element moves like the old one; very different motion can trigger a full regeneration. ### How many reference images can I use, and which aspect ratios are available? You can attach up to 5 reference images to lock a character, product or location, and say in the prompt what each one is for. Output is 16:9 or 9:16 only, so square and cinemascope framing are not available. ### What does Gemini Omni struggle with? It struggles with clips longer than 10 seconds, dense multi-action choreography, and close hand-and-prop interactions such as eating, where props can vanish or change. Packing too many actions into one clip can break character consistency, and strict content filters may block edgy or conflict-heavy prompts. ## Try Gemini Omni Video on Fuser Grounded, sound-on clips and plain-language edits that keep your character and scene intact from the first take to the last [Roll the shot](https://app.fuser.studio) ## Recipes using Gemini Omni Video - [Reframe Video without Reshooting](https://fuser.studio/recipes/social-media-video-resizer.md): Upload one source video to create vertical 9:16 and landscape 16:9 versions for social content. ## Used for - [Creative agencies](https://fuser.studio/solutions/industries/creative-agencies.md): Develop pitches, campaign assets and interactive client work in Fuser. Combine AI generation, editable design and reusable processes across your agency. - [Film & animation](https://fuser.studio/solutions/industries/film-animation.md): Develop film treatments, character and environment concepts, storyboards and motion studies. Connect image, video and 3D exploration in Fuser. - [Social content](https://fuser.studio/solutions/use-cases/social-content.md): Create social graphics, carousel artwork and short videos with a recognizable brand style. Keep copy editable and recurring content easier to produce in Fuser. - [Video & storyboards](https://fuser.studio/solutions/use-cases/video-storyboards.md): Develop storyboard frames, shot variations and video studies from your references. Compose clear boards and explore movement in one Fuser project. ## Related models - [Kling 4.0](https://fuser.studio/models/kling-4-0.md) - [MiniMax H3 Reframe](https://fuser.studio/models/minimax-h3-reframe.md) - [Kling 3.0 Video](https://fuser.studio/models/kling-3-0-video.md) - [MiniMax H3](https://fuser.studio/models/minimax-h3.md) - [Seedance 2](https://fuser.studio/models/seedance-2.md) - [Grok Imagine](https://fuser.studio/models/grok-imagine.md) Node reference: [Gemini Omni Video docs](https://docs.fuser.studio/docs/nodes/video/gemini-omni-video.md) ## More articles - [How to Edit a Video With Text Prompts: Kling O3 Edit, Gemini Omni, Grok and Luma Tested](https://fuser.studio/articles/edit-video-with-text-prompts.md) - [Best AI Video Editing Models: Same Clip, Same Edits](https://fuser.studio/articles/best-ai-video-editing-models.md) - [Best AI Video Generators With Audio: Same Dialogue Prompt, Eight Models](https://fuser.studio/articles/best-ai-video-generators-with-audio.md)