Gemini Omni Video Edit: AI Video Editor

Gemini Omni Video Edit — natively multimodal video-to-video editing. Restyle, relight, swap subjects, or rewrite scenes from a source clip with a single prompt. Unified reasoning across modalities preserves motion and audio continuity while applying the edit.

📝

Overview

About this model

Gemini Omni Video Edit brings Google's natively multimodal any-to-any model to source-driven editing. Provide a clip and a natural-language edit instruction — restyle the look, change the season, swap subjects, or rewrite dialogue — and the model rewrites the video in a single pass while preserving original motion and timing.

Because visuals and audio are reasoned about jointly, edits stay coherent across modalities: regenerated audio matches the new visuals, and original ambient sound can be preserved when desired.

1Restyle & relight: Convert a live-action clip into watercolor, anime, claymation, or any other visual style with a single prompt.
2Localization: Rewrite spoken dialogue into a new language while preserving the speaker's identity and lip motion.
3Continuity fixes: Adjust wardrobe, props, or background details that didn't make it on set, without reshooting.
4Creative iteration: Test multiple looks for the same shot — golden hour vs. neon night, summer vs. winter — from one source.
5Content adaptation: Reframe and restyle footage for new platforms without rerunning the original shoot.
💰

Pricing & Value

Cost analysis

muapi$2.40 (720p/1080p) · $3.60 (4K)

Flat rate per generation — same price regardless of duration. Synchronized audio included at no extra charge.

Fal.aiNot available

Gemini Omni Video Edit is not currently available on Fal.ai.

ReplicateNot available

Gemini Omni Video Edit is not currently available on Replicate.

* Competitor pricing is estimated based on similar model architectures and usage tiers.

⚙️

Technical Details

Configuration schema

Edit Promptstring

Describe the edit to apply. Examples: restyle to watercolor, change the season, swap the subject's outfit.

Default ValueTransform the growing clock into @image1 while preserving the original expansion and wall-breaking motion. As the clock enlarges, metallic spider legs unfold from the sides and stab through furniture and walls. Keep the camera motion and apartment destruction intact.
Reference Imagesarray

Reference images. 7 image slots available total. Video uses 2 slots; each character_id uses 1 slot. Max 20 MB per image.

Default Valuehttps://cdn.muapi.ai/assets/gemini-omni-video-edit.jpg
Source Videostring

Source video to edit (max 100 MB, max 30 s). Optional if image_urls are provided.

Default Valuehttps://cdn.muapi.ai/assets/gemini-omni-text-to-video.mp4
Trim Start (seconds)number

Start of the clip window to edit, in seconds (>= 0).

Default Value0
Trim End (seconds)number

End of the clip window to edit, in seconds. Must be within 10 s of trim_start.

Default Value10
Duration (seconds)Enum (4 options)

Duration of the generated video in seconds.

Default Value8
ResolutionEnum (3 options)

Output video resolution. 720p and 1080p are the same price; 4K costs more.

Default Value1080p
Aspect RatioEnum (2 options)

Output video aspect ratio.

Default Value16:9
Audio IDsarray

Up to 3 voice profile IDs returned by the Gemini Omni Audio endpoint.

Default Value-
Seedint

Random seed (0–2147483647). Fix for reproducibility; results may still vary due to model stochasticity.

Default Value0
Character IDsarray

Up to 3 character IDs from Gemini Omni Character to feature in the video.

Default Value-
📖

Implementation Guide

Developer documentation

How to Use Gemini Omni Video Edit

  1. Upload a source video Provide a public video_url or upload a clip from the playground. Shorter inputs (under ~15 seconds) produce the most consistent edits.

  2. Write a clear edit prompt Be specific about what should change and what should stay. Example: 'Restyle the entire clip as a hand-drawn Studio Ghibli animation, keep the original camera motion, lighting direction, and timing.'

  3. Choose resolution

    • 720p / 1080p — same price ($2.40 per generation)
    • 4K — higher resolution ($3.60 per generation)
  4. Submit and poll POST to /api/v1/gemini-omni-video-edit and poll GET /api/v1/predictions/{request_id}/result until status is completed.

  5. Prompt tips

    • Anchor what should not change: 'keep the original framing', 'preserve subject identity', 'maintain camera path'.
    • For style transfer, name the reference style explicitly: '1980s VHS', 'oil painting', 'claymation'.
    • For dialogue edits, quote the replacement line and note the desired voice tone or language.

Common Questions

Frequently asked

How is this different from running a text-to-video model with the source as reference?

Gemini Omni Video Edit conditions on the source clip directly and reasons across visuals and audio jointly. The model preserves motion, timing, and continuity from the source while applying the edit — instead of generating a new clip that loosely matches.

Can I edit just the audio (e.g., change dialogue) without changing visuals?

Yes — disable `preserve_audio` and write a prompt that targets only the audio (e.g., "replace the dialogue with: '...'"). The visuals will stay aligned with the source while the audio regenerates.

What video lengths are supported?

Maximum clip length will be confirmed at launch. We expect support for inputs up to 30 seconds, matching the text-to-video variant.

Will the edit affect motion or camera movement?

By default the model preserves the source clip's motion and camera path. If you want the camera or motion to change, say so explicitly in the prompt.

When will Gemini Omni Video Edit be available on muapi?

Google announced Gemini Omni at I/O 2026 on May 19, 2026, with API access rolling out in the following weeks. The video-edit variant will go live on muapi as soon as the upstream API supports it.