- AnFlowerBand2小时前
Hi 论坛,
We’re building Genjutsu AI, a browser-based AI video editor for reworking existing footage with reference photos.
The workflow is simple: upload a source clip, add photos of your character, pet, or product, and choose what you want to change. You can also start with a video template from the homepage.
There are two main modes:
Motion Transfer: Use a source clip’s movement, timing, and camera motion to guide a new character or scene. Useful for dance clips, performances, and trend remixes.
Object Swap: Replace a person, product, outfit, prop, or background while aiming to preserve the surrounding shot. Useful for product demos, fashion previews, and ad concepts.
Prompts are optional. Start with the default instructions, or add a description when you need to assign two performers or specify a replacement.
Before generating, you can choose between Lite, Standard, and Pro tiers and see the estimated credit cost. Lite and Standard output at 480p; Pro outputs at 720p. Finished videos can be previewed and downloaded as watermark-free MP4s. One-time credit packs and subscription plans are available.
There are limitations: complex occlusion, fast movement, and interactions between multiple subjects can produce artifacts. Small packaging text, logos, and clothing details may also change. We recommend testing a short clip first and reviewing the result before publishing.
Try it here: Genjutsu AI https://genjutsu.im/
Free trial credits: Eligible new users receive free credits after account verification. You’re welcome to give it a try—we’d love to hear what you make!
The homepage includes templates and source/result comparisons, so you can explore the examples before trying a generation.
We’d love feedback: would you use this more for character and trend remixes, or for product, fashion, and advertising concepts? Specific workflow suggestions and examples of where it falls short are welcome.
Please use footage, images, and likenesses that you own or have permission to use.
主楼 😯