The trend, explained

Hotel Lobby AI Video Trend: What It Is & How to Make One

The orange booth. The shared mic. The unexpected duo. Here’s where it started and how to make your own version.

Hotel Lobby AI Editorial team8 min read
Two fictional performers beside a suspended microphone in an orange recording studio
AI-created scene concept. Not a video generated by this tool.
On this page
  1. What is the trend?
  2. Origin & timeline
  3. How to make one
  4. Two creation methods
  5. Compare services & test results
  6. Copy the prompt
  7. Choose your photos
  8. Common problems
  9. Music & permissions
  10. Questions & sources

The quick answer

Hotel Lobby AI is a rap-duo video format inspired by Quavo and Takeoff’s orange-studio performance on A COLORS SHOW. Creators use AI to put new people or characters into a similar scene: two performers, one suspended microphone, and a vivid orange room. Some tools edit an existing performance; others generate a new one from photos and a prompt. The name comes from the song, not an actual hotel lobby.

Create video

What is the Hotel Lobby AI video trend?

You’ve probably seen the format while scrolling: a fixed camera, a completely orange room, and two unlikely performers sharing a mic. They nod, gesture, react to each other and take turns performing. What changes from one remix to the next is the duo.

Friends, partners, fictional characters and animals all appear in remixes. The visual setup is recognizable even when the subjects change, so viewers understand the joke quickly. Results differ by workflow: replacing characters in a reference clip and generating a brand-new performance are not the same task.

Where did the trend come from?

The visual reference is Quavo and Takeoff’s performance of “HOTEL LOBBY (Unc & Phew)” on A COLORS SHOW. The official performance uses a minimal orange set and a microphone suspended between the two rappers. Its simple framing makes the scene easy to recognize and reinterpret.

Watch the original A COLORS SHOW performance

The original performance

The original COLORS video established the orange-room, two-performer visual. The original performance is linked below so you can compare it with later remixes.

The AI remixes

The research behind this guide documents AI remixes circulating in September 2026. Know Your Meme collects reported early examples. Treat “an early documented example” as different from a proven first creator; the linked source is a record of reporting, not a complete history of every upload.

Read the documented trend history

Why does this format work?

A distinctive room gives every version a common visual language. An unexpected pairing makes the scene personal or funny. The consistent arrangement also makes it easy to compare different versions. You don’t need a complicated plot to understand who the new duo is or why they are together.

How to make a Hotel Lobby AI video

Start with two character references and decide whether you want a new orange-booth scene or an edit of an authorized performance. That choice determines what inputs and output settings the service needs.

Choose two clear photos

Use one image for each performer. Prefer visible faces, natural lighting and a simple background. Make sure you own the images or have permission to use them and the people shown.

Assign left and right roles

Upload one photo for each performer and choose who appears on the left and right. Swap positions without uploading both again. Your draft is saved on this device so you can return to it before generation.

Choose your video settings

Choose a 5-second or 10-second video, the frame shape and video quality. The generator shows the credit cost before submission. Your finished video uses the matching fixed soundtrack; exact lip synchronization is not guaranteed.

Generate, play and download

Sign in with Google and click Generate video. Follow the upload and generation progress beside the form. When your video is ready, play it on the right and download the MP4. Saved results are also available in My videos.

Create video

Two ways to create the Hotel Lobby effect

The tools can look similar on the surface while doing different jobs. Choose based on the result you want rather than the model name alone.

Reference-video character replacement

An existing clip provides the timing, framing and motion. The service replaces the performers with new subjects. This can stay closer to a specific choreography, but needs a suitable reference video and rights to the source material.

Best for a close match to a specific performance.

Original orange-booth generation

Photos and a prompt guide a new scene. You can ask for two performers, a single microphone and an orange background without copying exact gestures. Motion and facial consistency still depend on the model, and audio is a separate capability.

Best for an original variation of the visual format.
What mattersReference editingOriginal generation
Main inputAuthorized clip + character photosCharacter photos + prompt
MovementGuided by the source performanceGenerated anew, less exact
AudioMay preserve source audio; check rightsProvider-dependent or added separately

How to compare Hotel Lobby AI video services

Compare services that make the kind of video you want. A short, silent orange-booth generation and a longer edit that follows an existing performance are different products. Check actual examples showing both input photos and the finished clip, rather than judging quality from a promotional still.

Check what the price includes

Before paying, confirm duration, resolution, framing, audio and watermark conditions. If a service offers trial credits, check eligibility and export restrictions. Read how failed requests and retries are charged, and whether refunds restore credits or money. A low price per attempt can still mean a higher cost for a usable clip if you need several retries.

Make the comparison repeatable

Keep the same two authorized photos and use comparable settings. For a small comparison, try ten outputs per workflow and decide what counts as usable before reviewing them. Record every attempt, including failures, so the strongest result does not hide the cost of getting there.

  • Record the model and version, prompt, reference clip if used, duration, resolution and audio settings.
  • Check both identities, left/right positions, hands and motion. Record elapsed time, total cost and how many clips met your criteria.
  • For a failed result, change one photo or setting at a time. Keep the before-and-after clips and note what improved.

This is a testing checklist, not a benchmark of every listed service. Our two real generation tests are shown on the homepage.

A Hotel Lobby AI video prompt to start with

If you use a general image-to-video tool instead of a preset, use this prompt as a starting point. Attach the two character references and check whether that tool supports multiple subjects.

Orange studio rap duo

Create a cinematic two-person rap performance in a minimalist burnt-orange recording studio. Exactly two performers stand side by side with one microphone suspended between them. The left performer begins with rhythmic hand gestures and subtle head movements while the right performer reacts naturally, then takes a turn performing. Use a stable front-facing camera, soft studio lighting, consistent character identities, natural movement, and a seamless orange background. Keep both performers in their original left and right positions. Avoid additional characters, extra microphones, face merging, flickering, distorted hands, and abrupt camera movements.

Keep the subject count and left/right positions explicit. Start with a fixed camera and short duration. A prompt helps guide the model, but it cannot guarantee perfect identity preservation or synchronized lips.

Best photos for this trend

Good references reduce ambiguity. The model should be able to see each face clearly and distinguish one person from the other. A long description cannot recover facial details missing from a blurry source.

Photo conditionRecommendation
Clear, front-facing portraitA good starting point; slight angles can also work.
Even natural lightKeep the eyes and facial features visible.
Masks, strong shadows or blurReplace with a cleaner reference when possible.
Several people in one photoChoose separate images for the left and right roles.

Common problems and what to try

The two faces merge

Use distinct references, assign left and right roles explicitly, and avoid overlapping bodies. If the service does not support two separate character references, choose a workflow that does.

The characters switch sides

Ask for fixed left/right positions and fewer camera changes. Reduce gestures that move one person in front of the other. Reference-based editing can be a better fit when exact blocking matters.

Faces change between frames

Try a sharper photo, less motion and a shorter test. Compare whether one person consistently drifts more than the other. Replace that reference first rather than changing every setting at once.

Hands or arms look unnatural

Fast gestures and overlapping limbs are difficult for video models. Ask for subtle, natural movements, keep the camera stable, and review a short output before creating a longer clip.

The lips do not match the music

Check whether the workflow explicitly supports audio synchronization. Swapping the soundtrack does not automatically correct mouth movements. Silent video, generated sound and licensed music are different options.

The original song and your permissions

This site does not bundle the original song or performance footage. An AI edit does not itself provide rights to the soundtrack, source video or someone’s likeness. Before posting, check the permissions for your images, audio and publishing platform.

An original scene with your own photos and audio you have permission to use offers a different workflow from remaking the original performance. Keep synthetic content labeled where appropriate. This is a practical reminder, not a statement that any particular use is licensed.

Questions before you start

Sources & further reading

The original performance and the documented remix history are useful starting points. Product capabilities and costs vary by service and should be verified in the generator you actually use.

Your version starts with two photos.

Now that you know the format, prepare your own pair of performers in the generator.

Create video