Updated 2026-10-07
Hotel Lobby AI Prompts
Every Hotel Lobby AI prompt describes the same five things: a seamless matte burnt-orange studio, one black condenser microphone hanging from a thin cable at centre, two full-body performers fixed left and right, an alternating performance (one leads, the other reacts, then they switch), and a locked-off centred wide shot with no cuts or camera moves. The prompts below are written for general AI video tools; paste them into Kling, Veo, Seedance or any model that accepts reference photos.
Prompt 1 — Classic orange booth duo
Vertical 9:16 video. Two performers stand full-body inside a seamless matte burnt-orange studio booth, Performer A on the LEFT and Performer B on the RIGHT, with one black condenser microphone hanging from a thin cable at exact centre between them. Performer A leans in and raps first with restrained hand gestures while Performer B nods and reacts; then Performer B takes the lead while Performer A responds. Conversational timing, never mirrored. One continuous locked-off centred wide shot, soft even frontal light, hands and shoes visible. No cuts, zooms, pans, extra people, text, logos or watermarks.
Prompt 2 — Best friends
Same booth and framing as Prompt 1. Two adult friends in everyday streetwear; keep each person’s face, hair, complexion, outfit and height from the reference photos. Friend A opens with confident but natural hand movement; Friend B waits, listens and gives small independent reactions, then steps toward the mic and performs while Friend A nods back.
Prompt 3 — Couple duet
Same booth and framing. A couple, one on each side of the hanging mic, trading verses with playful eye contact. Keep both identities distinct and both bodies fully in frame. Alternate leads halfway through.
Prompt 4 — Two pets
Same booth and framing. Cat A upright on the LEFT and Cat B upright on the RIGHT, one microphone at centre. Cat A performs first with small realistic paw gestures and head rhythm while Cat B watches; then they switch. Preserve breed, markings, eye colour and collars. Avoid human faces, merged bodies, extra paws or duplicated tails.
Prompt 5 — Two babies
Same booth and framing, two toddlers in high-contrast outfits, exaggerated head nods instead of hand gestures, gentle alternating rhythm, everything else identical.
Prompt 6 — Same person twice
Same booth and framing. The SAME person appears on both sides in two different outfits taken from two different reference photos, arguing in verse with themselves. Keep the face identical on both sides; only clothing and posture differ.
Why prompting alone disappoints
Text describes the scene but not the identities. Without two separate reference photos and a motion reference, the model tends to average the two faces, swap sides mid-clip, or animate only one performer. That is why the working versions of this trend run on reference-to-video models — and why the Hotel Lobby AI generator here skips the prompt entirely: the booth and performance are preset, and your two photos are the only input. Full walkthrough on the how-to page.
FAQ
Which model works best with these prompts?
Models that accept multiple reference images and a reference video (Seedance 2.x, Kling 3 with motion control) keep the two identities apart. Text-only prompting on any model tends to merge the faces.
Can I skip the prompt?
Yes — the generator on this site has the booth, framing and motion preset; you only upload two photos.