Hotel Lobby AI trend
Upload two photos and our AI puts you both in the orange COLORS booth from the viral Hotel Lobby AI trend, trading verses at the mic. Same moves and camera cuts, with the Hotel Lobby song already in sync.
4:3 · 768p MP4 · no watermark
Only upload photos of people who agreed to appear in the video.
The Hotel Lobby song is already in the video and synced to the rapping — just download and post.
Examples
Made from two photos with Nana Banana Pro. Every clip has the Hotel Lobby song synced to the rapping, so turn the sound on.
The clip most people post: the verse at the mic, the point at the camera and the turn.
The full 31-second performance, from the close two-shot to the wide view of the orange booth.
Two friends trade the verse in the 15-second version.
The camera pulls back to the whole orange studio and the yellow floor, then cuts back in.
The trend
The Hotel Lobby AI trend, also called the Migos AI trend, takes Quavo and Takeoff's 2022 COLORS performance of "Hotel Lobby" and swaps in new faces. Since September 2026 people have used it to put themselves, friends and couples at the mic, then post the clip with the song on TikTok and Instagram Reels.
Despite the name, nobody is in a hotel. It's the orange COLORS studio: two performers, one hanging mic and a yellow floor.
Person 1 stands on the left in a striped shirt, Person 2 on the right in an orange polo. Both rap, point at the camera and trade verses.
No filming, template or editing app. The AI recasts every shot with your faces and keeps the moves and camera cuts.
How it works
Three steps and about two minutes of rendering.
One photo per person: Person 1 takes the left spot, Person 2 the right. Clear half-body photos work best.
The 15-second version is the clip most people post; the full take runs 31 seconds. Rendering usually takes about two minutes.
Save the MP4 with the Hotel Lobby song already synced, or switch the song off and add the sound in TikTok.
Photo tips
The AI keeps the face and hair from each photo, so better photos give a closer match.
Show your head, hair and shoulders. Tight face-only crops often come out less like you.
Crop out friends and busy backgrounds. Each spot at the mic needs its own photo.
A clear, front-facing photo works best. Sunglasses, masks and hands over your face hide what the AI needs.
Daylight or soft indoor light keeps detail. Very dark or heavily filtered photos lose it.
FAQ
Quick answers about the Hotel Lobby AI trend and this generator.
It's a viral AI video trend where people recreate Quavo and Takeoff's COLORS performance of "Hotel Lobby" with new faces. This generator recasts that performance with the two people in your photos.
Upload a photo for each person, choose the 15-second or full version and generate. When it's ready, download the video and post it — the song is already in it.
Signing up is free, but each video uses credits: the 15-second version costs 30 credits and the full 31-second take costs 50. If a render fails, the credits are refunded automatically.
No. The trend is named after the song. The scene is the orange COLORS booth where Quavo and Takeoff performed it, and that's the scene your video recreates.
No. The Hotel Lobby scene is built in, so all you need is two photos. There's no template to download and no video to upload.
Yes, by default. The song from that part of the performance is added and synced to the rapping. To add the sound in TikTok or Reels instead, switch the song off before generating and you'll get a silent video.
Each video recasts both performers, so it needs two photos. To appear twice, upload two different photos of yourself.
Choose 15 seconds or the full 31-second take. You get a 4:3, 768p MP4 with no watermark, with the Hotel Lobby song unless you switch it off.
Videos are made with MiniMax H3. Each shot is recast separately, keeping the faces from your photos and the moves, timing and camera cuts of the original performance.
No. Nana Banana Pro is an independent AI video tool and isn't affiliated with Quavo, Takeoff, Migos, COLORS or their labels. Only upload photos of people who agreed to appear in the video.
Two photos, one mic, the whole performance in about two minutes.