Genjutsu AI

Genjutsu AI is an independent, browser-based video transformation workspace that uses the Higgsfield Genjutsu model to turn existing footage into new scenes, characters, or objects. Positioned as a focused video-to-video editing tool rather than a text-to-video generator, it requires a source clip and uses 1–8 reference images plus a text prompt to guide changes. Its core features include two main modes: Motion Transfer, which reimagines a performance or scene around the source video's movement, and Object Swap, which replaces a selected product, prop, outfit, or character within an existing shot. Users can trim clips to 4–30 seconds, crop framing, choose 480p or 720p output, and preview credit costs before generating. The target audience includes content creators, social media marketers, filmmakers, editors, and AI enthusiasts who want to iterate on video ideas without a subscription. Content features include community examples, source-and-result comparisons, a tutorial, a gallery of templates, and a case study. User experience emphasizes transparent costs, private uploads, MP4 downloads, and no auto-renewal. Technical features include support for MP4, MOV, and WebM uploads up to 500 MiB, upload preparation as H.264 MP4 up to 45 MiB, browser-based decoding and exporting, private job history, and temporary storage of source clips for 7 days and completed results for 30 days. Pricing is credit-based with one-time packs and no subscription.