Why is it called hotel lobby when there is no lobby?
Because the name belongs to a song, not a room. The frame everyone copies comes from a 2022 studio performance of Hotel Lobby by Quavo and Takeoff: two performers on opposite sides of one hanging microphone against a bright orange wall, one rapping while the other reacts, then switching. When people search "hotel lobby AI" and expect marble floors, that is the mix-up. The trend is three fixed parts: the orange backdrop, the single mic, and two performers taking turns. Tools like casttake rebuild only those three parts and drop your two faces into them, which is why you see the same booth across thousands of unrelated pairs.
Can I put my cat in it?
This is the most reasonable question, since the first AI versions that spread in September 2026 were two cats. The answer depends on how the tool works. Template tools that keep a face from a photo generally want human faces, because face keeping is the whole mechanism. casttake is explicit about it: the photo check looks for a real face, and pets, cartoon characters and faces photographed off a screen are turned away. That is a real limitation for a meme that started with cats.
The workaround it suggests is sensible. For a pet or character duo, use a general image to video tool and describe the scene in the prompt: an orange studio backdrop, one hanging microphone, two performers taking turns. You lose the guaranteed face matching, but cats were never going to need it. If your only photo shows a person holding a pet, crop the pet out and upload the person alone.
Which photo goes on the left, and does it matter?
It matters more than people expect. In a template like this, the first photo you add goes on the left of the screen and the second on the right. The left performer opens the first verse while the right one nods along, then they swap and finish together at the mic. So put whoever should lead on the left. If the clip comes out with the wrong person opening, there is nothing to fix in the photos themselves. Add the same two in the opposite order and render again.
A related habit worth building: watch the first two seconds before sharing. Face swapping between sides is the problem people most often report with this format across tools, even where a given template has not shown it.
Will TikTok mute it for copyrighted audio?
Not if the tool generates its own audio. The versions made with casttake use made up rap lines, a beat and voices produced for the template, plus a booth and mic built from scratch. No original recording, no original footage, no artist faces. That keeps the clip about you and your friend rather than a copy of someone else's recording, and it leaves a platform's audio matcher nothing to catch.
The audio is not the only rule, though. The clip is a vertical 9:16 MP4, so it fits TikTok, Reels and Shorts, but both people in it must have agreed to their face being used, and the template asks you to confirm that before you upload. Photos of public figures or of anyone who has not said yes are off the table, and they are the uploads that get videos taken down.
Top comments (0)