About Genjutsu AI
An independent reference-video creation tool.
What we build
Genjutsu AI helps creators explore a different character, setting, or visual style from a short reference video. The current workflow uses MiniMax H3 Max Reference through fal.
What we offer today
One reference video, an optional image, and a prompt produce a new generated clip. You can choose 5, 10, or 15 seconds, 480P or 768P, and landscape or portrait. Account history, background task processing, and video downloads are available.
Our relationship to Higgsfield
We are independent of Higgsfield and do not run its proprietary Genjutsu model. Similar creative use cases do not imply identical quality or control. Exact source timing and pixel-preserving object replacement are not promised.
Early access
Pricing, paid subscriptions, and checkout are deferred. Generation uses provisioned account credits. Contact support for access or questions.