
Good starting point
- One subject in each photo.
- Face visible, looking roughly toward the camera.
- Even light, with natural skin and facial detail.
- Include the upper body and any outfit you want to reference.
Two photos. One Hotel Lobby rap duo video.Upload two photos to create a Hotel Lobby rap duo video, with the reference performance guiding the scene, movement, camera and timing.
Hotel Lobby AI puts a new duo into a rap performance inspired by Quavo and Takeoff's 2022 A COLORS SHOW appearance.
The orange backdrop, hanging microphone, and two-person setup make the scene recognizable. Our template uses a 15-second performance reference: Photo 1 guides the performer who starts on the left, and Photo 2 guides the performer who starts on the right.
Watch the original COLORS performanceThe clip here shows the reference scene. Your photos guide the new cast; the reference guides the movement and camera. Generated visuals and audio can differ from the source.
The original performance provides the context; your two photos provide the cast.
COLORS published the HOTEL LOBBY performance by Quavo and Takeoff, also known as Unc & Phew, in June 2022.
Watch the original COLORS performanceThe compact set keeps attention on the pair, their gestures, and their reactions. It gives a two-photo remake a clear role for each subject.
Upload two separate portraits, review the 15-second reference and credit price, then generate and inspect your version.
Three generated examples. Compare the faces, gestures, and outfits before making your own.
Choose your duo, check the reference and price, then review the finished video.
Photo 1 guides the performer who starts on the left; Photo 2 guides the performer who starts on the right. Use one clear subject in each photo.
Preview the Hotel Lobby reference. The current template makes a 15-second, 16:9 video at 720p. Sign in and review the credit total before submitting.
Play the full result and check both faces, gestures, and audio. Download the MP4 from your video history when you are happy with the result.
Use one photo for the opening left performer and one for the opening right performer. Show the outfit if you want it to guide the edit.


Each choice in the workflow is designed around maintaining two separate subjects through one fast, shared performance.
Photo 1 is assigned to the opening left performer and Photo 2 to the opening right performer. The role mapping stays tied to identity when they move.
The model uses visible clothing, colors, fabric details, footwear, and accessories from the photos. If an outfit matters, show it clearly in a half-body or full-body image.
The Hotel Lobby video guides the choreography, camera, cuts and timing. The output includes audio, which may differ from the source; preview the complete result before sharing.
Upload, preview, pay, monitor, and download from the browser. Saved task recovery helps prevent duplicate paid submissions after a refresh.
Compare this ready-made two-person scene with a CapCut template or setting up an AI video model yourself.
The current 15-second Hotel Lobby template costs 300 credits. Review the total in the tool before submitting. Failed generation tasks restore their reserved credits.
Keep the two-person idea, or try a single-subject dance.
Cast a keyboard player and a partner in the white-studio scene.
Explore generatorGive one pet a dance reference, with photo tips for animal subjects.
Explore generatorTry a cat-focused dance with a photo and motion reference.
Explore generatorUpload one photo for each of two distinct performers. The photos guide their appearance, while the source video guides the scene, movement, camera and timing. The result includes audio, but its sound and visuals may differ from the reference.
Hotel Lobby is the built-in two-performer reference shown in the generator. Upload two photos and the video guides the scene, movement, camera and timing. Audio is included, but may differ from the reference.
The photos guide identity, but generative video can still change facial details, hands, clothing, or lip sync. Use sharp source photos and review the complete result before publishing it.
Use photos you own or are authorized to use. Do not impersonate public figures, deceive viewers, harass anyone, or suggest a real person endorsed words they never said.
AI video queues vary and a render can take several minutes. The task is saved to your account, so reloading the page does not create a second paid request.
The task system restores the credits reserved for a failed or canceled generation. A completed provider file is copied to private app storage before it is offered for download.
This workflow expects two separate images so each subject can be assigned to a specific performer. Crop a group photo into two individual images before uploading, and keep one clear subject in each file.
The current template requests a 720p, 15-second landscape video in 16:9. A separate 9:16 output option is not currently available.
Custom lyrics and voice controls are not available in this workflow. The video provides a performance reference, and the model generates an audio track that may differ from the source.
You can experiment with authorized pet or character images. Results are less predictable when the subject's anatomy differs from the two human performers, so inspect faces, paws, limbs, clothing, and fast motion before sharing.
Choose two photos and bring your duo to the orange studio.