Building fun AI video tools has been an exciting journey. Today I want to share RapSong, my new SaaS project inspired by the viral Hotel Lobby rap trend.
Lots of people love that orange recording booth duet style, but there was no simple workflow to let two people from static photos perform alternate rap lines. Existing AI music tools can generate songs, and AI video tools can animate images, yet combining lyric writing, beat production, lip sync and character identity preservation in one single pipeline is tricky.
This tool is built for casual entertaining videos: birthday roasts, wedding surprise clips, class reunion performances, corporate team shout-outs and even funny pet rap duets. No camera shooting, no video editing skills required.
The biggest technical challenge I faced is maintaining consistent facial features during animation, avoiding flickering, and synchronizing lip movements to the generated rap audio. I keep tuning the model pipeline to improve stability and visual quality.
If you want to try it out, visit RapSong and create your own custom rap duet video from two photos. I’d love to hear your feedback. What kind of fun AI-generated video would you make first?