Hotel Lobby AI Prompt (Copy and Paste)

Copy-and-paste Hotel Lobby AI prompt

This prompt is based on the one our own generator uses in production. We refined it over dozens of test videos until the orange walls and the hanging microphone came out right nearly every time. Attach one photo per performer as reference images, in left-to-right order, then paste:

Vertical 9:16 music performance video, about 8 seconds. Image 1 is the LEFT performer and Image 2 is the RIGHT performer; these are two separate people, not first and last frames. Keep both faces and identities recognizable. The two performers stand side by side in a vivid, evenly lit orange recording booth; every visible wall is orange. One black studio microphone hangs on a thin cable from the top center of the frame, between them, with its bottom unattached. No microphone stand, no boom arm, no tabletop microphone, no furniture. Medium upper-body framing, static camera. Both perform with natural rap gestures and rhythmic energy for the whole clip. Original upbeat instrumental hip-hop beat. No existing song, lyrics, logos or watermarks. No extra people, no merged faces.

Video models differ, so treat this as a starting point. If your tool has a separate “negative prompt” box, move the “no microphone stand, no furniture” part there.

What each part of the prompt controls

Every line in the Hotel Lobby AI prompt is there because something went wrong without it. The drawing below maps the main parts of the prompt onto the scene.

What we learned testing the prompt

Our first version of the prompt only described an orange booth and a microphone. In early tests, some videos drifted to pale backgrounds, and some swapped the hanging microphone for a microphone stand. Saying “every visible wall is orange” and spelling out what the microphone must not be fixed both problems.

With the final Hotel Lobby AI prompt, 42 of 44 test videos passed our frame-by-frame review. The two misses were about motion, not the scene: in one, both performers walked in from the edges during the first second; in the other, a performer turned away from the camera briefly.

Choosing reference photos for your prompt

With image-to-video models, the reference photos matter more than the wording. A perfect Hotel Lobby AI prompt cannot rescue a dark, blurry or side-on photo. Give each performer a sharp, front-facing, head-and-shoulders picture with even light, one subject per image. Photos with similar lighting help the two performers look like they share the same room. Crop out other people, text and stickers before you upload.

Prompt variations that work

Keep the core of the prompt the same and change one line at a time. These tweaks came out of our own testing:

• For pets, name each animal and its side: “Image 1 is a ginger cat on the LEFT, Image 2 is a beagle on the RIGHT.” Add “each animal keeps its natural size and sits upright” to stop a cat being drawn at human size or standing on its hind legs.
• For a person and a pet, say which one is the person: “Image 1 is a woman on the LEFT, Image 2 is her dog on the RIGHT.”
• For more energy, change “natural rap gestures” to “big, confident rap gestures and head nods on every beat”.
• For a calmer, deadpan clip, ask for “minimal movement, slow nods, serious expressions”.

Common prompt problems and fixes

• Only one performer appears: add “both appear together for the entire clip”.
• Faces drift or blend: use sharper, front-facing reference photos rather than a longer prompt.
• A microphone stand appears: add “no microphone stand, no boom arm, no tabletop microphone”.
• The background is not orange: repeat that every visible wall is orange, and ban beige, gray and yellow.
• Performers walk in or out of frame: add “both performers are in place from the first frame, static camera”.
• The model adds a real song or lyrics: ask for an instrumental beat and say no lyrics.

Check the result before you post

This is the checklist we use when we review a video made with our Hotel Lobby AI prompt. If any line fails, change one thing and try again:

• Both faces are recognizable, and each performer looks like their photo.
• Photo 1 is on the left and Photo 2 is on the right for the whole clip.
• Exactly two performers: no extra people, merged faces or extra hands.
• The microphone hangs from above, between them, with no stand or desk.
• Every wall is orange, from the first frame to the last.
• The audio is an original beat, not a real song with lyrics.

Skip the prompt entirely

Writing and testing a Hotel Lobby AI prompt takes time, and every failed try on a paid model costs money. My Hotel Lobby AI runs the tested production prompt and settings for you. You only choose two photos and get an 8-second, 9:16 vertical MP4 in 720p for $5.99, with no prompt writing, model picking or editing, and your credits come back automatically if a generation fails.

Frequently asked questions