One skill, many jobs

Editing modes

Inputs

MP4 or MOV · up to 200 MB

This preset requires a replacement image. Use one clear unobstructed character with body coverage similar to the clip. A face-only photo leaves body details to generation. Reference only the character, not the background.

Generation

Model

3–15.5s video, up to 4 reference images

Resolution · Generate audio

This replaces the character’s wider appearance; use Face Swap for facial identity only. For multiple people, identify the target by position and original clothing. The person preset retains original clothing; creative characters adopt target clothing. Edit either instruction as needed. Inspect turns, hands, held objects and occlusions. Illustrations are not submitted as references.

Presets

Credits update automatically with your inputs.

Result

A new look for your footage

Upload a clip, describe your changes and choose a model

Upload videoMP4 or MOV · up to 200 MB
Edit instructionDescribe your changes and add optional reference images
Generate editGenerate and view your result here

Your edits

Your edits will appear here.

credits ↗

Video re-enactment

Recast any motion with your characters, locations and products or swap specific elements and keep the rest

Character replacement console

Same engine as /video-to-video/character-replacement with the Character Variation preset selected.

Total control, zero reshoots

  • Keep only the motion

    The motion, camera, and timing stay the same, everything else is rebuilt from your references

  • Keep your character in every video

    Shoot on your phone, Genjutsu does the production: wardrobe, set, grade

  • Same take, two realities

    Drag the handle to compare the original and the new version. The motion and cuts stay in sync, only the scene changes

How to re-enact a video

  1. 1Upload your MP4/MOV source (clear subject, stable lighting).
  2. 2Add a reference image for the new character and pick Character Replacement.
  3. 3Describe what to swap, confirm the credit quote, and generate.

Genjutsu FAQ

What is the reference effect Genjutsu?

Genjutsu lets you transform an existing video in two ways: transfer its motion into a new cast, location or look, or swap selected elements while keeping the rest of the shot intact.

What's the difference between Motion Transfer and Object Swap?

Motion Transfer keeps the motion, camera and timing while rebuilding the scene from your references. Object Swap changes a selected character, outfit, product, location or object while preserving the rest of the original shot.

What can I change with Genjutsu?

You can recast characters, change locations, introduce products or wardrobe, or replace individual elements in an existing video using your own references and prompts.

How long can my reference video be?

Genjutsu currently supports reference videos from 4-30 seconds, as shown in the generator.

How many reference images can I add?

You can add up to 30 reference images for characters, products, wardrobe and other visual elements.

Do I need video editing experience?

No. Upload your video and references, choose Motion Transfer or Object Swap, describe what you want to change, and generate.

How does pricing work?

Generations use the standard the reference effect credit system. Each run shows its credit cost before you generate, so you always know what a render will use.

Can I use the results commercially?

Yes — what you generate is yours to publish across organic and paid channels, subject to the reference effect's Terms of Use.

Character ReplacementObject Edit

Re-enact your footage

New cast, same performance.

Open character replacement