How to Make a Hotel Lobby AI Video

Turn two adult portraits into a duet inspired by a reference performance. The portraits identify the left and right performers; the reference supplies movement and timing. This guide explains the four-step workflow and what to check before sharing a result.

Public generation and purchases are not open yet: reference permissions and model validation must be completed first. This public preview does not accept sign-in or photo uploads; the steps below describe the planned workflow. No generation credits are consumed while the service is unavailable. The initial scope is duets using portraits of adults who have agreed to take part.

1. Sign in and choose two original photos

Open the generator, sign in and upload one clear portrait for each person. The first photo is assigned to the left performer and the second to the right. You do not need to combine them into a collage. Follow the photo guide for file limits, lighting and unobstructed faces, then check both previews.

2. Choose duration, resolution and aspect ratio

Planned durations are 5, 10, 15, 20, 25 and 30 seconds, with 480P, 720P and 1080P delivery, portrait 9:16 or landscape 16:9. Choose Hotel Lobby or a recording studio background; the reference motion remains the same. These specifications are under validation and cannot currently be ordered. Final delivery and any upscaling will be disclosed before generation opens.

Review the video options before submitting. A shorter duration is useful for checking how your two portraits behave in the performance, without committing to a longer clip.

3. Review credits, permissions and audio

Check the displayed credit cost for your settings and confirm permission to use both photos; reference footage and audio permissions are the operator’s responsibility. See pricing for the available purchase information.

Keep original soundtrack is on by default. The site automatically combines the reference clip's audio with the generated picture. No separate extraction or editing tool is needed. Turn it off for a silent MP4. This removes audio only; the reference motion stays the same and the price for the same settings is unchanged. Missing audio or a track shorter than the selected duration is flagged before submission, without consuming credits; choose a shorter duration or disable the soundtrack.

4. Preview both faces, listen and download

After a task succeeds, preview the video and download its MP4. Check each person's identity and side throughout the clip, then listen to the audio and inspect mouth timing. Identity, backgrounds and movement may vary. Original example portraits illustrate inputs only; no matching verified video is offered on this website yet. Lip sync may differ from the recording.

Follow your task after leaving the page

Tasks are saved to your account, so you can return to My videos to check progress and view completed results. Submission, queueing, processing and saving are active stages, with no fixed completion-time promise.

A confirmed generation failure restores the deducted credits once. An uncertain submission needs review and is not automatically resubmitted. Visual dissatisfaction alone does not qualify as a confirmed technical failure; read the refund policy. Review the privacy policy before uploading personal photos.