Choose your plan

Standard videos are 480p, based on 12 seconds each. Shorter videos let you create more. The quote shown before generation determines the actual credit cost.

Genjutsu AI is a video-to-video editing approach that starts with existing footage instead of generating every shot from scratch. A source video provides the motion, timing and camera movement, while reference images guide how characters or other visual elements should change. The broader Genjutsu workflow is commonly associated with motion transfer and targeted element replacement. AI Genjutsu currently starts with character replacement: upload a source video and a reference character to create a new version of the same performance.

Upload a source video and a reference character image, describe the person you want to replace, review the generation cost and create the edited video.

Yes. Character replacement is the current AI Genjutsu workflow. Short clips with one clearly visible person usually provide the best starting point.

The source video guides the body movement, timing and camera path of the edit. Results can vary depending on motion, occlusion, source quality and the reference image.

Motion transfer uses movement and camera behavior from an existing video as a guide for creating a visually different version of the shot. AI Genjutsu currently starts with character replacement rather than exposing a separate full motion-transfer mode.

Short continuous clips with one clearly visible subject usually work best. Heavy occlusion, rapid cuts and multiple overlapping people can make consistency more difficult.

Use a clear image where the character's face, hair and overall appearance are easy to see. Similar framing to the source subject can also help.

No. Video generation requires credits because AI video processing has significant compute costs. You can explore the workflow before purchasing credits.

No. You can buy credits separately without a subscription. Monthly and yearly plans are also available.

No. AI Genjutsu is an independent AI video editing service and is not affiliated with or endorsed by Higgsfield.

Start with an existing performance, add a visual reference and create a new version while keeping the motion and camera direction you already like. Change the on-screen character while following the original movement, timing and camera path. Use a different reference character to create a new visual version of the original shot.