

Turn two photos into an AI rap duet video for TikTok, Reels and Shorts.
AI rap duo video generator: add two selfies, pick a stage and get a 12-second rap video of you and a friend, with a beat and a verse about your names. No app needed.

It's a TikTok and Instagram Reels trend, sometimes called the orange booth trend or the AI rap duo trend, in which people use AI to put themselves and a friend or relative into a live-studio rap clip. Two performers stand side by side in an orange booth and rap at one microphone. This page makes your own Hotel Lobby video from your photos, with nothing to film or record.
In the maker above, add two selfies or one photo of you both, pick the stage and type who the song is for. Sign in, check the credit cost, then confirm. When it's ready, download the MP4 from your history and post it.
No. Everything runs in your web browser on a phone or computer, so there's nothing to install. There's no CapCut template to download and no prompt to write: the stage, the beat and the performance are already set, and you only type the names and the occasion for the verse.
Yes. Each clip comes with an AI-made beat and two AI-generated voices rapping a short original verse. It is not the song the trend is named after and does not copy any real artist or voice. You can still swap the audio for a sound the platform offers when you post.
Yes. Under "Who's the song for?" type the names and the occasion, up to 120 characters, and we write a short original verse around them. The left person raps the first line, the right person answers, and both finish together. Leave it empty for a general hype verse. You can't paste in a full song of your own.
Yes. Choose One photo of both and upload a photo where both faces are clear and side by side; the person on the left raps first. Two separate selfies usually keep faces closer to the photos, so use them when you have them. Both people must be adults who agreed to appear in the video.
Four: the orange booth from the original trend, a luxury hotel lobby with marble, gold and chandeliers, a dark studio booth with purple neon, and a street cypher at night with a graffiti wall. Every stage is generated; the trend's name comes from a song title.
The standard video is about 12 seconds, a vertical 9:16 MP4 at 720p, which fits TikTok, Instagram Reels and YouTube Shorts. You can choose a shorter or longer clip, a lower or higher resolution, or a square or widescreen format; the exact length and credit cost appear before you confirm. Many platforms ask you to label AI-generated content, so use that label when it applies.
No, it isn't free. The standard video costs 10 credits, and other models, resolutions and lengths cost more or less; the cost is always shown before you confirm. Credits come in one-time packs that never expire. If a confirmed generation fails, the credits charged for that attempt are returned once. You can request a refund of unused paid credits within 168 hours of payment.
Not of a celebrity or public figure. You can use photos of other people only if you own the photos or have permission, and only if everyone pictured is an adult (18+) who agreed to be included. Photos of minors are not allowed, and you shouldn't make a video of anyone without their consent.
No. It doesn't paste faces onto existing footage from the trend. A new scene, with picture and sound, is generated from your photos in one pass. That's why faces, hands, clothing and movement can vary, and why exact likeness is not guaranteed.
A confirmed failure, including a result blocked by the safety check, returns the credits charged for that attempt once. If your connection drops while you submit, use the recover option instead of starting over; it checks the same request without a second charge. If a finished clip looks off, try sharper, front-facing photos. Each new attempt uses credits.
Looking for something different? See Framecomet alternatives