Hotel Lobby AI Trend: How to Make the Orange Booth Video

The Hotel Lobby AI trend is not set in a hotel. It is two performers, one hanging microphone, and a flat orange wall — and since mid-September 2026 it has been one of the most-copied AI video formats on TikTok and X.

Cats have done it. TV characters have done it. Couples, best friends, and brand mascots have done it. This guide covers where the format comes from, the two ways to make one, and the exact prompt language that keeps the orange booth looking like the orange booth.

What is the Hotel Lobby AI trend?

It is an AI video format that places two new subjects — people, pets, or characters — into the staging of a 2022 music performance: side by side at a single microphone hanging from the ceiling, in front of a saturated orange backdrop, trading verses in one locked-off shot. Most versions are made from just two photos.

Where the orange video comes from

The source is real, and it is worth knowing before you remix it.

• May 2022: Quavo and Takeoff of Migos release the single "Hotel Lobby" as the duo Unc & Phew — Quavo is Takeoff's uncle
• June 2022: They perform it on A COLORS SHOW, the Berlin series known for filming artists against one flat color. Theirs is orange. The video has since passed 100 million views on YouTube
• November 2022: Takeoff dies, months after the session was filmed
• September 2026: A TikTok edit swapping the pair for two characters from the TV series Suits takes off, and the format spreads to hundreds of thousands of versions

Quavo responded by reposting the untouched performance and calling it "The Original," as Complex reported. The trend has also drawn criticism — over Takeoff's legacy, and over real people's faces being used without their involvement. That shapes how to do this well, covered below.

Why this format works

• The frame never changes. One camera, no cuts, no movement. Viewers recognize it in half a second
• Two subjects create a story. One performs, one reacts, then they swap. The pairing is the joke
• The set is one color. A flat orange wall is the easiest environment an AI model can hold consistently
• The entry cost is two photos. No editing skill required for the basic version

Two ways to make a Hotel Lobby AI video

Route 1: a template swap tool

Dedicated trend tools take two photos and map them onto the original performance — choreography, camera, and timing come from the source clip, so the result syncs to the track. Fast, and identical to everyone else's. You control the cast and nothing else.

Route 2: prompt it yourself

Build the scene from scratch in a video model such as Seedance or Kling. Slower, but you control outfits, gestures, framing, clip length, and who reacts when — and you end up with an original performance rather than a copy of someone else's. This is the route that stands out in a feed full of identical swaps, and the one the rest of this guide covers.

The prompt workflow, step by step

Step 1: Prepare two clean reference photos

One subject per photo. Front-facing, well lit, upper body visible, plain background. Sunglasses, hats, and group shots are where identity drift starts.

Step 2: Generate the still frame first

Lock the composition as an image before you animate anything. In an image model that accepts multiple references, such as Nano Banana Pro or GPT Image:

Vertical 9:16 studio photograph. Two performers standing side by side facing camera, framed from the knees up. The person from reference image 1 on the left, the person from reference image 2 on the right. A single vintage condenser microphone hangs from the ceiling on a black cable, centered between them at mouth height. Seamless saturated orange backdrop and matching orange floor, no visible horizon line. Soft even frontal studio lighting, minimal shadows. Left performer in a black leather jacket, right performer in a cream varsity jacket. Locked-off camera at chest height, 35mm lens.

Step 3: Animate the still

Feed that frame to your video model as the start image and describe the performance beat by beat:

Static locked-off shot, no camera movement, no cuts. Left performer leans toward the hanging microphone and raps with confident hand gestures, head bobbing on the beat. Right performer nods along, points at the left performer, adds short ad-lib reactions. At the midpoint they switch: right performer steps to the microphone and raps, left performer steps back, arms crossed, nodding. Orange backdrop stays flat and unchanged. Both faces remain consistent with the start frame.

Step 4: Add the sound inside the app

Export the clip silent and add "Hotel Lobby" from the TikTok or Instagram sound library when you post. That keeps the audio licensed and groups your clip with the rest of the trend.

What to lock in every prompt

• Backdrop: "seamless saturated orange backdrop and floor" — not just "orange background," which invites gradients and props
• Microphone: hanging from the ceiling, centered, mouth height. A mic stand breaks the look instantly
• Camera: static, locked-off, no zoom. State what the camera does, not what it should avoid
• Positions: assign left and right by reference image, every time
• Roles: one performs while the other reacts. Two subjects rapping at once reads as noise
• Wardrobe contrast: different jackets or colors keep the two subjects from blending into each other

Variations that are getting views

• Pets: two cats or a cat and a dog in tiny streetwear — different fur colors keep them readable
• You and your best friend, partner, or sibling — the format's most shareable version
• Original characters: your own mascot, game character, or illustrated duo
• Brand versions: two products or two team members, for shops and creators riding the trend
• Color flips: same staging, your brand color instead of orange — recognizable, but yours

How to do it without the backlash

• Use faces you have the right to use. Yourself, consenting friends, pets, or original characters. Skip celebrity and politician swaps
• Do not recreate Quavo or Takeoff themselves. The format is the staging, not their likeness
• Turn on the AI-generated label when you post. Platforms expect it for realistic AI video
• Credit the original. Naming Quavo, Takeoff, and COLORS in the caption costs nothing and keeps the source visible

Mistakes that break the look

• Camera drift. Any push-in or pan and it stops reading as the trend
• Busy backdrop. Studio lights, logos, or furniture in frame. The wall is empty
• Animating from text only. Without a locked start frame, faces drift within seconds
• Too much in one clip. Keep each generation to one exchange; stitch longer versions in an editor

Build the full prompt set in minutes

The trend rewards people who post early and post variations — which means writing the still-frame prompt, the motion prompt, and the reference instructions again for every new duo. The Seedance Prompt Builder turns that into 32 dropdown fields and generates five outputs per concept, including a reference file brief for character lock and a camera movement brief — so a new pairing takes a few clicks, not a rewrite. Working in Kling instead? The Kling 3.0 Prompt Builder covers Motion Control and Elements for the same two-subject setup.

Hotel Lobby AI trend FAQ

What is the Hotel Lobby AI trend?

An AI video format that puts two new subjects into the staging of Quavo and Takeoff's 2022 COLORS performance of "Hotel Lobby": an orange backdrop, one hanging microphone, and two performers trading verses.

Why is it called Hotel Lobby?

The name comes from the song, not the setting. There is no hotel in the video — only an orange studio.

What is the original orange video?

Quavo and Takeoff's "Hotel Lobby" session on A COLORS SHOW, published in June 2022.

How many photos do I need?

Two — one clear, front-facing photo for each performer.

Which AI tools can make it?

Template swap tools handle the one-click version. For a custom version, use an image model that accepts two references for the still frame, then a video model such as Seedance or Kling to animate it.

Can I use the real song?

Add it from the TikTok or Instagram sound library when posting rather than baking it into the file, so the audio stays licensed on the platform.

Pairs well with

• Nano Banana Pro Prompt Builder — for the two-reference still frame the whole clip is built on
• Character Prompt Builder — for creating an original duo instead of using real faces
• AI Video Prompt Builder Pro — for running the same concept across every major video model

Related guide: Character Consistency in MidJourney: Fixing Panel Drift.

Browse the full AI Video Prompt Builders collection.