Two photos in · Prompt already written · Sound arrives with the picture
Hotel Lobby AI from two uploads — you and a friend under one mic in a bright orange booth, the duo clip your friends will replay all week.
Four pairs, one booth. Every clip here was posted in the last ten days, and every one of them keeps the same bones: orange wall, one hanging mic, two people who look like they own the room.
A groom in a white tuxedo and a bride in lace, lifted out of their wedding album and dropped under the mic. The finger-to-lips hush at the top of the verse is the beat everyone copies, and it lands harder in a wedding dress.
Put the two of us inA black dress on one side, a pink corset on the other, both of them performing to the lens. Vertical, short and made for the feed — this is the cut that fills the trend on reels.
Make one with my friendTwo men who would never share a stage anywhere else, both completely committed to the bit. A square crop keeps them tight on the mic and ready for any feed.
Pair me with anyoneA man in a linen suit and a woman in white, both behind dark glasses, trading lines like a headline act. Dress the photos up and the booth dresses up with them.
Try it with our photosA Hotel Lobby AI video puts two people into one famous setup: a bare orange booth, a single microphone hanging between them, two performers trading a verse straight to camera. It started from Quavo and Takeoff's COLORS performance of Hotel Lobby, and within days thousands of people were posting their own pairs — couples, siblings, best friends, a parent and a kid. The Hotel Lobby AI version is that setup built around your two photos.
You bring two photos. The model brings the booth, the lighting, the hanging mic, the hand movements and the steady camera. Where most template apps paste new faces over the original footage, this page generates the whole scene fresh — the same way the dance tool builds motion out of a still — so what comes back belongs to you and your partner.
Creative engine
Both photos go in as references, one per performer. The prompt carries the booth, the mic, the moves and the framing. Length and resolution are yours, and the credit estimate updates before anything runs.
One of the most-shared versions pairs a man with a horned warrior. If it has a face in a photo, it can take a side of the mic.
A son and his mom, trading bars like they have done it for years. The less likely the duo, the harder people rewatch.
Making a Hotel Lobby AI video takes two photos and one click. The part people spend days on — telling the model which face belongs to which performer — is already written into the prompt on this page.
Face the camera, good light, head and shoulders clear. The first photo becomes the performer on the left, the second the one on the right, and a sharp photo is what keeps each face locked to its side of the mic.
The orange wall, the hanging mic and the straight-on framing are already in the prompt. Add one line if you want matching suits, a wedding look or streetwear, and leave the rest alone.
Give it enough seconds for both of you to take a turn — the handoff is the whole clip. Sound comes back with the picture, so the clip is ready to post.
The pairs getting the most replies are the ones nobody expected to see rapping together. Pick yours before the feed moves on.
The version everyone posts first. Two faces the people who follow you already know, one mic, and a performance neither of you ever gave. Two all-black outfits make the booth look like a record label shoot.
Mom and son, dad and daughter. The generational mismatch is the punchline, and it is the kind of clip that ends up in the family group chat within the hour.
A dog or a cat takes the other side of the mic and bobs along to your verse. It is the pairing nobody else in the feed thought to post, and the one people send to everyone they know.
One real face, one invented one — a mascot, an OC, a creature. Start from a drawing rather than a photo with animate illustrations.
The questions people are actually asking in the threads where Hotel Lobby AI clips are being made and argued about.
Here it is two photos in and a new booth performance out, with no source video to find, trim or face-swap. Most of the clips going around were made the long way: the original performance video cut to length and fed to a model with two photos, plus a close-up and a full-body shot of each person, then several attempts to get the right face on the right side.
No. Template apps that paste your faces over the original music video can get stopped by copyright checks after you have already paid. Nothing from the original is uploaded or reused on this page — the booth, the moves and the camera are generated fresh around your photos — and the credit cost is shown on screen before anything runs.
That is the most common way this goes wrong: the model is never told clearly which photo is which performer. Upload one person per photo, in the order you want them left to right, and keep the prompt's lines that assign the first photo to the left performer and the second to the right. Clear, front-facing photos do the rest.
Yes. One photo of you, one of your dog, and the dog takes the other side of the mic. Animals read best when the photo shows the whole head clearly and front-on, and the same goes for a cat, a drawing or a mascot.
No. A lot of people feel uneasy seeing the original duo turned into a template, and this page is built so their performance is never touched: no frame of the original video goes in, and nobody appears in the clip except the two performers from your uploads. What you post is your pair, in a booth of your own.
The sound on your clip is generated with the picture, not taken from the original track. Most people add the song itself when they post, from the sound library in TikTok or Instagram, and line the hush gesture up with the drop.
Credits are shared across everything Kavel runs. You see the cost before each run, and a failed job is never charged.
Pay safely and securely with
Payments are processed by Stripe over an encrypted checkout. Your card details are handled by Stripe and are never stored by Kavel.
The Hotel Lobby AI wave is peaking now, and the pairs getting watched are the ones posted while it still is. One photo of you, one of whoever takes the other side of the mic, the prompt already filled in, and the credit estimate on screen before anything runs.
