Hotel Lobby AI Video Generator

It composes an original two-person performance scene staged in a minimal orange studio booth with a suspended microphone and warm booth lighting. You supply one photo for each person and an optional direction note; the generator turns them into a short silent video with original movement and framing.

The two photos are sent as ordered subject references to the video model: the first photo stands for Person A and the second for Person B, and both guide one shared scene. They are not the first and last frames of the video, so neither photo has to look like how the clip starts or ends.

Yes. The note is a creative suggestion of up to 300 characters — for example a camera framing or the mood of a synchronized step. It is optional, and it never guarantees a specific action, composition or camera move in the result.

The current model generates 9:16 or 16:9 video at 480p or 720p, for 4 to 30 seconds. The settings only list values the selected model actually supports, so an unavailable combination cannot be submitted.

The estimate uses the same shared credit function as the create request, based on the selected model, duration and resolution, and it updates whenever a setting changes. If a free trial is still available, the estimate for that creation is zero.

No. Videos are silent by default and there is no built-in music control. If you need a soundtrack, add audio you have the rights to after downloading the MP4.

No. The photos guide each character's look in the scene, but the model does not guarantee face identity, likeness, lip-sync or exact poses. Treat the result as an original stylized performance, not a portrait reproduction.

No. Generation is prompt-led from your two photos and note only; there is no video upload, no reference-clip copying and no attempt to reproduce an existing music video, dance routine or film scene.