Hotel Lobby AI: Make an Orange Booth Duo Video with MiniMax H3

The Hotel Lobby AI look is an orange booth performance, not a hotel lobby.

• “Hotel Lobby” is a 2022 song, not a lobby set. The performance people refer to is Quavo and Takeoff on A COLORS SHOW, uploaded to YouTube on June 17, 2022.
• Public descriptions put two subjects in a seamless orange booth, taking turns, usually vertical and about 10 to 15 seconds, with one hanging microphone between them.
• Complex reported the format gained traction in early September 2026 after videos of two cats.
• Public clips often reuse the original performance. This page does not. Use only your own photos or photos of consenting adults, never public figures or minors.

Text to video needs no photo. Image to video starts from one still of both people.

• On the homepage, Orange booth rap duo is text to video at 9:16. The Use this prompt button fills in the text below. No photo is required.

Two [young friends in streetwear] stand side by side in a small, bright solid-orange studio booth, each rapping into a vintage silver microphone on a stand, like a live music session. They trade lines, nod to the beat and point at each other. Locked-off medium shot, warm even studio light. A punchy hip-hop beat with two voices trading rap lines about [their best-friend road trip].

• Replace the square brackets with your own people and your own topic. Do not name a public figure or a minor.
• Image to video uses one picture of both people in the orange booth as the first frame, plus a prompt for motion and sound. A last frame is optional. Image generation is not offered.
• A task can use a first frame, a last frame, or both. It cannot place two separate portraits side by side. Video-to-video and video editing are not offered.

The ready prompt breaks into people, place, action, camera, and sound.

• People: Two [young friends in streetwear]. Keep the brackets to people you may depict.
• Place: a small, bright solid-orange studio booth.
• Action: side by side, each rapping into a vintage silver microphone on a stand, then they trade lines, nod, and point. The public look often uses one hanging microphone; this prompt uses a stand for each person. You can edit that line. The prompt limit is 7,000 characters.
• Camera: Locked-off medium shot, warm even studio light. The camera stays fixed.
• Sound: a punchy hip-hop beat with two voices trading rap lines about [their best-friend road trip]. Every H3 video includes sound. Leave music or dialogue out and the model picks the sound, often quiet ambience. Use 9:16, or 21:9, 16:9, 4:3, 1:1, or 3:4. Image to video follows the uploaded frame.

A clip is 4–15 seconds at 768P or 2K, and some requests are refused.

• Pick a whole number of seconds from 4 to 15, and pick 768P or 2K. Usually 2–5 minutes.
• 768P costs 11 credits per second, so 6 seconds at 768P is about 66 credits. 2K uses more credits per second than 768P.
• Credits are reserved when a task is accepted and released automatically if generation fails. A video that is generated but blocked by our final safety check is still charged. Eligible videos can be submitted for manual review within 7 days.
• This site cannot put you into the original video, recreate that choreography, lip-sync to that performance, or use the original song.
• Prompts and uploads are screened. Sexual content, minors, a likeness without consent, and infringing content are not allowed. Upload JPEG, PNG, or WebP, up to 5 MB, 256–5760 px per side, aspect ratio 0.4 to 2.5.

Common questions