Everything About Higgsfield Genjutsu

Genjutsu is Higgsfield’s video editing model, launched on August 31, 2026. It rebuilds part of an existing video, such as a person, a setting, an object, or the whole style, while keeping the original motion, timing, and camera work. It works on real footage and AI-generated video.

Shoot once. Change the world around the motion as many times as you like.

Motion Transfer

Applies the movement from your video to a new character or rebuilt scene

Object Swap

Replaces one named element, such as a product, outfit, or prop, and keeps the rest

Restyle

Added on September 30, 2026. Redraws the whole video in an animation style

Restyle runs inside Motion Transfer as a styles option. See our Genjutsu Restyle guide for details.

Source video

One clip, 4 to 30 seconds

Reference images

Up to 30 per generation in the web app, for characters, objects, outfits, products, locations, or styles

Prompt

Optional. A preset or short description is enough

Output

480p, 720p, or 1080p

Genjutsu has more than 30 presets. A preset alone can drive the change, and a short description adds specific direction.

Higgsfield listed these costs for a 15-second clip, verified in September 2026. Both modes cost the same:

480p

40 credits, about $2.00

720p

104 credits, about $5.20

1080p

144 credits, about $7.20

On the Higgsfield API, Genjutsu was listed from $0.159 per second at a 50% discount in mid-October 2026. Prices change, so check before you generate.

• The Higgsfield web app, under video tools
• The AI Influencer Motion tab, to give your characters real movement
• Through the Higgsfield MCP connector in AI assistants
• The Higgsfield API, for developers. See our Genjutsu API guide

• Choose Motion Transfer or Object Swap
• Upload a 4 to 30 second source video
• Add reference images if needed
• Pick a preset or write a short description
• Generate, compare with the source, and export

• Turn anime into live-action, or live-action into anime
• Try many outfits on the same clip
• Make ad versions for different markets with the same pacing
• Turn raw phone footage into a polished look
• Give AI influencers natural movement
• Swap products in demo videos

• Use steady, well-lit footage
• Keep the subject clearly visible and not too small
• Use clear reference images from several angles
• Start with short clips at 480p while testing
• Move to 1080p for the final version

• Only edit videos you own or have permission to use
• Get consent from people who appear in the video
• Use only faces you have permission to use
• Never make real people appear to do things they did not
• Label AI-edited video where platforms require it