Hotel Lobby AI

Workflow guide · updated October 5, 2026

What does “Hotel Lobby AI swap” actually change?

A Hotel Lobby AI swap can mean putting yourself in an orange-booth performance or replacing a face inside an existing clip. Our tool does the first: it generates a new video from one or two photos. It does not accept the original performance video or perform a frame-by-frame face replacement.

New photo performance

Input: one Solo photo or two Duet photos. The model invents motion from a prepared reference and directions. Timing, expression and composition can differ from the original.

Swap inside a source video

Input: an existing video and replacement identities, with a workflow designed for face replacement or motion reference. Check support for two identities, occlusion, audio and source-video rights before choosing a tool.

If you want to put yourself in the orange booth

  1. Choose a Solo or Duet template and upload photos you have permission to process.
  2. Start with 5 seconds at 720p, costing 50 credits. Add a small gesture rather than promising a copy of the entire original routine.
  3. Review both identities and the whole scene. A technically completed render can still have altered features, extra limbs or separate-looking compositions.
  4. Download the MP4. Add permitted audio separately if needed; a beat-aligned edit does not make mouth movements match the lyrics.

Can I make a free Hotel Lobby swap here?

The signup trial covers one new 5-second 720p photo performance with 50 welcome credits. It does not cover an original-video face-swap service, which this generator does not provide. Prompts are free; further renders consume credits. Read the exact trial conditions.

Which workflow should I choose?

Choose photo generation when the orange-studio look and recognizable subjects matter more than reproducing specific frames. If the original movement and timing are essential, look for a tool whose documented inputs include a source video. Our workflow comparison separates documented capabilities from our own rendering tests.