Put yourself in a viral clip with one photo — no prompt
I shipped Genjutsu AI — a small web tool for a narrow job: take a video people already share, swap the people for photos you pick, and keep the original take. No prompt. No storyboard. You upload the clip, the app splits it into scenes and lists who’s in frame. You add one still photo per person you want to change. Everyone else, the background, and the props stay as filmed. What stays the same…
Introducing Genjutsu AI, a web-based tool that allows users to swap people in viral videos with their own photos without any prompt or storyboard required. The application takes the uploaded clip, splits it into scenes, and identifies the faces in each frame. Users can then choose to replace any of the faces with their own photo while keeping the original shot order, timing, camera cuts, pacing, and audio intact.
To use Genjutsu AI, users simply upload their video, select a template (such as Hotel Lobby, Tiny Desk AI, Night Dance, or Bulin 47), and attach a photo for each face they want to change. Every scene in the video is automatically detected, and users can generate the new video with their chosen identity. The final product is a 720p video with the original audio and can be downloaded.
The underlying technology behind Genjutsu AI consists of a Next.js app with a Python cast pipeline for scene and face processing, Prisma + Postgres for managing accounts and credits, object storage for handling uploads, and a video model API for the swapping functionality. Users can sign in using their Google account, and all generations are stored on their account with credits not expiring.
Some important limitations to note when using Genjutsu AI include the maximum scene length of 15 seconds, the output resolution of 720p, and the 10 credits cost per 15 seconds of video. A failed run will refund the used credits. Users can sign up with Google, and their generated videos will be stored on their account. The first 15-second video on the Starter plan costs $6.90, while the Plus and Creator packs offer additional features on the pricing page.
The product is not the same as Higgsfield Genjutsu; Genjutsu AI requires the user to provide the video, while Higgsfield Genjutsu handles the video part as well. Users are encouraged to try Genjutsu AI, provide feedback on scene splits, face consistency, and credit pricing, especially if they create short-form or meme content.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
