How to make a Hotel Lobby AI video.

1. Choose one clear photo per performer

You don’t need to take the photos together. Pick one image for each subject and use the clearest version you have. People, pets and character images all use the same two upload slots; the model’s interpretation can vary.

• Keep one main subject in each image. A group photo makes it harder to establish who belongs in each role.
• Show the face and upper body. Leave room around the head and shoulders rather than cropping very tightly.
• Prefer sharp, evenly lit images. Heavy blur, strong beauty filters and objects covering the face give the model less detail to work with.
• Use JPG, PNG or WebP, up to 10 MB each. Convert HEIC photos to a supported format before uploading.

2. Upload and check the left and right roles

On the Hotel Lobby AI generator, Photo 1 guides the performer on the left and Photo 2 guides the performer on the right. Check each thumbnail. Replace an image if it looks wrong, or select Swap photos to change the order.

Your selected photos stay on this device until you submit a generation. A temporary local draft keeps your photos and settings ready for 30 minutes while you sign in or buy credits. Return to the studio to review them before generating.

3. Create your orange-studio duo

Sign in with Google and buy video credits. Choose 5, 10 or 15 seconds and either 768P or 2K. Choose Size: 9:16 portrait (default), 16:9 landscape or 1:1 square, then select Generate video. The credit cost updates before you submit. Credit packs are one-time purchases.

The built-in Hotel Lobby AI prompt works with our fixed example video to guide the orange background, suspended microphone, gestures and camera timing. You don’t need to upload a reference clip, provide a soundtrack or write instructions.

The task moves through queued and generating states. Processing time varies with demand. You can leave the page and return to My videos without starting a new request.

4. Review the whole video, then download

Watch the result from beginning to end. Check both subjects’ appearance, the left and right placement, and the movement. AI can alter details, so a clear input photo helps without guaranteeing an exact likeness.

Use Download video to save the finished result. There’s no extra credit cost to download the same video. Videos may become unavailable later; save the ones you want to keep.

The output uses the Size you select; the existing example videos are landscape. If you want a particular song, add audio in an editor or the social platform where you post, using music available to you there. This tool does not promise the original recording.

If something needs a fix

The file won’t upload

Check its format and size. Renaming a file extension does not convert the image. Export it as JPG, PNG or WebP, then select it again.

The task is still pending

Keep the existing task. A slow queue or a temporary status-check error is not a confirmed failure. Your credits remain reserved while we check the provider’s result.

The generation failed

A confirmed failure restores the reserved credits used for that request. Choose Try again after the task is marked failed. A completed video you simply want to remake uses a new credit allowance.

The two subjects look blended or change appearance

Try separate, uncluttered images with one clearly visible subject per photo. Avoid tiny faces and strongly different lighting. A new attempt is a new generation and uses the credits shown for your selected duration and quality.