How Do I Make a Hotel Lobby AI Video?

Quick answer

To make a Hotel Lobby AI video, open Duet Booth, sign in, and upload one clear photo for each performer. Choose Original moves for the Hotel Lobby reference, or New performance for a fresh duet. Check the available settings and credit cost, buy credits if needed, then generate. Open History to check progress and download your finished video. Availability and results vary; the tool does not promise an exact song or movement match.

What do I need before making my first video?

Prepare two separate photos: one for the left performer and one for the right. Each image must be a JPEG, PNG, or WebP file no larger than 10 MB. Use photos you have permission to upload and animate. A friend or pet can be your second subject; both subjects do not need to have been photographed together.

Choose photos with visible faces and enough detail to recognize each person or pet. Use the photo checklist below before uploading.

You also need to sign in to the account that will own the task and credits. The workspace shows your credit balance when it can retrieve it. A balance-loading error is not proof that your credits disappeared; refresh before buying again.

How do I upload the left and right photos?

Open the Hotel Lobby AI generator, choose the Create tab, and select Left performer. Pick the first photo, then repeat with Right performer. Check the two previews before continuing. Remove and replace the relevant image if you put a subject on the wrong side.

This is a real local workspace capture with no account signed in. The unavailable option shown in the capture cannot be submitted; your workspace displays its current availability and cost.

Do not upload the same group photo into both slots and expect the app to identify the intended pair. Each slot is designed for one visible subject. Cropping a group picture may help isolate a subject, provided the resulting face remains clear and you have permission to use the photo.

Which photos should I upload?

• One performer per photo: use a separate image for each person or pet. Avoid group shots with several possible subjects, and check the left and right previews.
• Portrait or full-body: choose a clear, well-lit portrait for facial detail. For clothing and body references, use a head-to-toe photo with the legs and shoes visible. The result may change clothing, facial features or composition.
• Clean, sharp images: avoid motion blur, dark shadows, heavy filters and covered faces. Remove status bars, captions, stickers and app controls from screenshots without cutting off the subject. Use JPEG, PNG or WebP files up to 10 MB each, and photos you have permission to animate.

These checks help make your intended pair clear; they do not guarantee likeness or successful generation. See the visual photo examples on the Hotel Lobby AI page for clear and difficult inputs.

Should I choose Original moves or New performance?

For the Hotel Lobby workflow, choose Original moves on the Hotel Lobby AI page. This mode uses the configured Hotel Lobby reference video with your two images. Its stage is fixed; changing the people does not mean you can select an unrelated stage for that reference.

Choose New performance when you want a fresh rap duet. This mode provides stage choices such as Hotel Lobby, Luxury Lobby, Studio Booth, and Street Cypher, plus an optional theme. A birthday theme describes the occasion; it is not a guarantee of exact lyrics or precise choreography.

The musical reference and a generated example are different things. The official performance is titled “Quavo & Takeoff - HOTEL LOBBY | A COLORS SHOW.” See the COLORS performance on YouTube. Its presence in a guide does not mean the generated file contains that recording or that you receive rights to reuse it.

How do I choose duration, resolution, and format?

Expand Video settings. The workspace offers 5, 10, or 15 seconds and landscape 16:9 or portrait 9:16. Landscape is the default. Use portrait when a tall composition is important; use landscape when you want more space around both performers.

Original moves lists 480P, 720P, and 1080P; New performance lists 480P and 720P. A listed setting is not necessarily available to submit: the server supplies the current availability and cost. Read Cost per video after changing a setting. If it says Available soon, choose an enabled option or return later.

For an initial experiment, a shorter, lower-resolution video lets you inspect the composition before spending more credits on another version. Higher resolution adds pixels; it does not guarantee a more accurate face or better performance.

How do I buy credits and start generation?

Use Pricing to compare a one-time credit pack with a monthly plan. Check the displayed price, credit amount, renewal terms, and expiry terms before choosing. The number of videos depends on the credit cost of each selected mode and setting; a credit pack is not a fixed count of every video type.

Continue through the checkout offered for your signed-in account. Buying credits and generating a video are separate actions: after purchase, return to the generator, check your balance, and press Generate video. Do not assume payment alone submits the photos.

When submission succeeds, the workspace displays the current task. While a task is active, the create button indicates that a video is in progress. If confirmation fails, read the error and use the available retry action; avoid starting repeated purchases or unrelated submissions just because you have not seen a result yet.

Can I leave the page while the video generates?

Yes. The current-task panel explicitly says you can leave and find the video in History later. Sign back into the same account, open History, and use Refresh History if the list needs updating.

Progress can move through waiting, generating, and saving. “Saving your video” means the app is still finalizing the result; it is not the same as a downloadable finished file. Completion time varies, so an elapsed timer is not a delivery promise.

If the page cannot refresh, it preserves the current task and offers a retry. Check the existing task before submitting a replacement. A connection error in your browser does not by itself tell you whether the generation failed.

How do I download and share the finished video?

When the task says Your video is ready, inspect the result in its player and choose Download video. Save the file on your device before sharing it. The player also has a Refresh video link action if its temporary playback link stops loading; the saved task remains the place to retrieve the result.

Review both subjects, the audio, and any unexpected details before posting. Explain that the performance is AI-generated when needed, and respect the subjects’ consent. A convincing-looking result should not be used to mislead someone about what a real person said or did.

What should I do if the result fails or looks wrong?

For a failed task, read the task error and credit status in the workspace. It distinguishes credits returned, no credits consumed, and a refund status still being checked. Confirm that state before paying for or submitting another attempt.

For a completed video with weak likeness, revisit the source photos: avoid blurred faces, heavy obstruction, and multiple subjects in one slot. A generation may still vary with clear photos. Download and judge the actual result rather than treating a promotional example as a guaranteed outcome.

Sources and further reading

• Hotel Lobby AI generator: current upload rules, modes, settings, availability, task status, and download controls.
• Duet Booth pricing: current purchase and credit terms.
• Hotel Lobby AI tool: the reference-based workflow and background of the trend.
• Hotel Lobby photo examples: clear portraits, full-body references and difficult inputs.
• Quavo & Takeoff — HOTEL LOBBY, A COLORS SHOW: the original performance used as background context, rather than a Duet Booth output.