Hotel Lobby AI Video Prompts: A Structure That Works

Template apps do not need a prompt, because they already hold the original motion and audio. Prompts matter when you use a general image-to-video model and want to rebuild the orange-booth scene yourself. The prompts below are written for this site and follow one rule: say what must stay fixed, then what changes.

The five parts of a good prompt

Format: vertical 9:16, about 15 seconds, locked camera, no cuts.
Setting: a plain saturated orange wall filling the frame, even studio light, one microphone hanging at the center between the two subjects, no props or on-screen text.
Identity: tell the model which reference image is the left subject and which is the right one, and ask for faces that stay consistent for the whole clip.
Motion: describe who leans toward the microphone, when they swap, and what the other person does meanwhile.
Constraints: no camera movement, no merged faces, natural hands, no duplicate limbs.

A two-person template

Vertical 9:16 video, 15 seconds, static waist-up camera, no cuts. A solid orange studio wall with one hanging microphone at the center. Use reference image 1 for the person on the left and reference image 2 for the person on the right. Keep both faces, hair and clothing consistent. The left person leans toward the microphone first while the right person nods and points at the camera. Around the halfway point they swap: the right person leans in while the left steps back and bobs. Both lean in together for the final beat and hold a small smile. No zoom, no camera movement, no warped hands, no merged faces.

Variations

Pet and owner: replace one identity with the pet, and ask for the animal to stand upright in a human-like pose. LightX notes that its template does this automatically for animals.
Solo: use one subject centered by the microphone. Some creators post solo versions, per LightX.
Character pairings: for fictional characters, say the style you want, such as live-action or cartoon, so the model does not blend them.

Audio and lip sync

Most general video models cannot legally or reliably reproduce the original recording, so results with prompt-only tools often have generic audio. Many creators add the sound afterward in an editor. Check the rights that apply to the audio you add. See the copyright and safety guide. This site does not publish the song's lyrics.

For more on getting two people to look distinct, read the two-person prompt guide. For a wider list of prompt phrases, see best prompts. Tools are compared on the tools page.

Practical points

Use this topic as part of the wider Hotel Lobby AI workflow: identify the original reference, decide what changes, generate a short test, inspect faces and motion, then check consent, labeling and platform rules.

Separate the 2022 source performance from newer AI-generated clips.
Use clear reference images and consistent framing.
Inspect hands, faces, lip movement and identity drift.
Use current platform disclosure controls where applicable.
Do not use private people's images without appropriate permission.

Follow Us