One image for the left performer and one for the right.
Hotel Lobby AI
Create a 15-second Hotel Lobby-style duet from two photos. Start with the editable default prompt, adjust the performance or dialogue, and generate a video with sound.
More video models
Explore other video models for your next project.
Your characters take the microphone
Create a 15-second Hotel Lobby-style duet from two photos. Start with the editable default prompt, adjust the performance or dialogue, and generate a video with sound.
A fixed duet format, with 768P or 2K quality.
To add dialogue or original lyrics, write short quoted lines in the prompt, specify the language and assign each line to the left or right performer. Words and lip sync may vary.
Choose the pair; the effect sets the performance
The prompt is filled in for you. Keep it or edit the setting, gestures and sound; use Restore default to return to the starting version.
Keep the left and right roles clear
Use one sharp image of a single character in each slot. The effect assigns the first image to the left performer and the second to the right.
Start with the default, then make it yours
To add dialogue or original lyrics, write short quoted lines in the prompt, specify the language and assign each line to the left or right performer. Words and lip sync may vary.
Choose quality and review the cost
MiniMax H3 powers this effect. Choose 768P or 2K and check the credit estimate before generating. Sign-in and sufficient credits are required; each run is charged separately.
One studio, different kinds of duet
A duet with your own lines
Create a 15-second Hotel Lobby-style duet from two photos. Start with the editable default prompt, adjust the performance or dialogue, and generate a video with sound.
Choose separate front-facing or slight three-quarter portraits. Avoid tiny faces, heavy blur and group photos. Keep visible details you want the model to preserve.
To add dialogue or original lyrics, write short quoted lines in the prompt, specify the language and assign each line to the left or right performer. Words and lip sync may vary.
Check the faces, mouth movements, sound and transitions before sharing. The preset keeps a consistent creative direction, but does not guarantee identical faces, exact words or perfect lip sync.

A playful character performance
The prompt is filled in for you. Keep it or edit the setting, gestures and sound; use Restore default to return to the starting version.
Give each character its own slot. References guide appearance, but stylized bodies, paws and outfits may change during animation.
The prompt is filled in for you. Keep it or edit the setting, gestures and sound; use Restore default to return to the starting version.
If a character is hard to recognize, try a clearer portrait. If the vocal exchange feels rushed, shorten the optional lines. Each new run uses credits.

How to make a Hotel Lobby AI video
Upload two character photos
Add one JPEG, PNG or WebP for the left performer and one for the right. Both photos are required.
Review the prompt and settings
Keep or edit the prefilled prompt, choose quality and aspect ratio, then review the credit estimate.
Generate and review
The preview shows the provided video until generation succeeds, then your result replaces it. Play the result with sound, check the pair and download it.
Hotel Lobby AI questions
What do I need to upload?
Upload one image for each performer. A reference video and default prompt are provided. You can edit the prompt directly; no separate video or music upload is needed.
Can I add lyrics or dialogue?
To add dialogue or original lyrics, write short quoted lines in the prompt, specify the language and assign each line to the left or right performer. Words and lip sync may vary.
Will it use the original Hotel Lobby song?
This effect does not promise a copy of the original song or a specific recording. The default directs an original upbeat groove with wordless vocals. Optional lyrics or dialogue can customize the vocal content.
Can I change the model or duration?
This effect uses MiniMax H3 with a fixed 15-second duration. Choose an aspect ratio and 768P or 2K quality; the default ratio is 4:3. For other models, durations and reference workflows, use the video model pages.
Will the two faces stay identical to my photos?
References guide appearance, but faces, clothing and gestures can change. Use clear individual images and review the entire clip for identity mixing, distorted hands and mouth movements.
Is this effect free?
Generation requires sign-in and SceneFlare credits. The estimate includes the reference-video input and selected quality.
Make the next duet your own
Upload one photo for each performer and choose the format. Keep the default duet or add a few lines of your own.




