How to make a Hotel Lobby AI video

What is the Hotel Lobby AI workflow?

Create a new performance from your photos and written direction. A seamless orange studio, hanging microphone and alternating gestures establish the scene. The original performance video is not supplied to the model.

1. Prepare your photos

Use one clear photo for Solo or two for Duet and Pet duet. Keep faces, hair and visible clothing clear. Upload the left performer first and the right performer second.

• One intended subject per photo.
• Avoid blur and hands covering the face.
• Use photos you have permission to use.

2. Choose the cast and scene

Choose Duet for two people or Pet duet for two pets. The detailed preset describes the orange studio, hanging microphone, stable framing and a newly choreographed performance. You can change the scene, lighting, camera and energy.

3. Choose 5 or 10 seconds

Hotel Lobby uses Seedance 2.0 Fast with 480p or 720p output, subject to account access. Choose 5 or 10 seconds to match the locally bundled soundtrack. The default ratio is 16:9.

Generation requires paid credits through a subscription or credit pack. Free trial credits alone do not unlock this workflow. Review current pricing before submitting.

4. Add an optional direction

The preset already describes opening head nods, alternating lead and reaction, small gestures below the face and a confident closing pose. Your note can change outfits, camera or movement.

5. Review and download with sound

Watch both faces, hands and outfits throughout the generated clip. The original soundtrack is added locally after generation, using the opening segment that matches your duration. It is not sent to the video model and does not guarantee precise lip-sync. Listen to the downloaded MP4 before sharing.

Do I need a source video or editor?

No source video is required. Photos and the detailed prompt create the new performance. The export pipeline adds the bundled audio. An editor is optional for captions or further trimming.

Why can faces or mouth timing vary?

The video is newly generated. Clear, separate portraits and modest motion help keep identities stable. The added recording is separate from model generation, so mouth timing can differ. Changing a photo or prompt starts another generation and uses credits.