AI video trend
Hotel Lobby AI Video Generator
Two people in a burnt-orange booth, no cuts — and none of it filmed. Felo's image-to-video seats your photos in the booth.
Open the video generator
Video generation is available to Pro users. Credits are charged by clip length and quality.
Sample made with Felo: one prompt, no editing
On X right now
The Hotel Lobby clips people keep reposting
Real posts from X, credited to their authors. Press one and it plays in X's player, likes and views attached.
Counts and links as of October 12, 2026.
The trend
What the Hotel Lobby AI video is
A talk-show segment that never existed: two people trading lines into a camera that never moves, in an all-orange studio.
It started as a real music video
The orange booth comes from Quavo's "Hotel Lobby" visuals. Once the look was recognisable, people started putting their own two names in those chairs.
One static camera, or nothing
A locked-off, full-body wide shot: no zoom, no pan, no cuts. Add movement and it stops looking like a studio segment.
The hard part is two faces at once
Left stays left, right stays right, and neither face drifts into the other. That is why the references get numbered in the prompt.
The artists, guests and shows people drop into this set are other people's names and likenesses. Use your own two people for anything commercial.
How it works
How to make your own Hotel Lobby AI video
No timeline, no keyframes, no editing suite.
01
Add one photo per person
Two clear, front-facing photos: one for the left seat, one for the right. Label which is which in the prompt.
02
Add a reference clip and the prompt
Upload a clip of the booth shot whose rhythm you want, then paste the prompt below. The clip carries the framing; the photos carry the faces.
03
Generate and download
Choose 5, 10 or 15 seconds and 768P or 2K, and keep the frame horizontal, 16:9 — a broadcast segment, not phone video.
Prompt
The Hotel Lobby prompt, ready to copy
Written in English, which the model reads best. Paste it in and swap in your own two.
Two-person booth shot
A burnt-orange studio with an orange floor curving up into the backdrop. Exactly two performers stand side by side, full body, feet visible, one on the left and one on the right. A single vintage silver microphone hangs from the ceiling on a thin cable between them. They trade verses: one mouths the words with confident head bobs and small hand gestures while the other nods and reacts, then they switch. Locked-off, static, front-facing, full-body wide shot — one continuous shot, no cuts, no zoom, no pan, no tilt. Photoreal, sharp faces, natural skin texture, warm even studio light. No text, no logos, no watermarks, no extra people, no duplicates.
Extra lines to append if the generation drifts
Identity lock, appended: "Keep each person's facial identity locked in every frame. The first reference image stays on the LEFT and the second stays on the RIGHT."
Cleanup, appended: "Exactly two subjects only. No additional people, reflections, duplicates or background characters."
Camera lock, appended: "Locked-off camera, fixed viewpoint, no zoom, no pan, no tilt, no camera movement."
Made with Felo
How Felo handles a shot like this
Nothing hides in this shot, which is exactly what Felo's image-to-video is built for.
Photos for identity, a clip for motion
Felo's image-to-video takes up to nine images plus reference video and audio, up to twelve files in one request. Two people stay two people instead of melting into one face.
5, 10 or 15 seconds, 768P or 2K
Tune the two references at 768P, then switch to 2K for the one you publish; 2K costs more credits per second.
One segment, or a whole series
Wrap the same references in an Agent that repeats the booth shot for every new guest or episode.
Generate your own version
Video generation is available to Pro users. Credits are charged by clip length and quality.
Questions about the Hotel Lobby AI video
Upload one clear photo per person, add a reference clip of the booth shot, paste the prompt above and generate. Felo renders both people in one generation.
Number the references in the prompt — image one is left, image two is right — and append the cleanup line so the model adds no third person or duplicate.
Horizontal, 16:9, with a camera that does not move. The locked-off wide shot is what makes it read as a studio segment.
Render time follows clip length and resolution: 5 seconds at 768P is the fastest combination, and 2K or longer clips take more.
Video generation is open to Pro users and spends credits by clip length and quality: 2K is 138 credits per second, 768P is 86. Every account also gets a daily free credit allowance.
More video trends
The generator itself, plus the other trends on the site.
Make your own Hotel Lobby video
The prompt is above and the generator is one click away. Bring your own two people.
Start generatingAvailable to Pro users.