Hotel Lobby AI Video Generator

Follow the example from inputs to download

The example above uses the displayed portraits and an 8-second reference. Its actual output is about 4.5 seconds at 768P, in 9:16. It is an existing generated sample with model audio, not a newly generated 8-second result.

1. 1. Preview the built-in clip, then add your left and right portraits. Keep the intended roles in that order.
2. 2. For a new 8-second result, select 768P and 9:16. An 8-second reference plus an 8-second output costs 128 credits. Keep reference audio is selected by default.
3. 3. Confirm your material permissions and generate. Open My Videos to check progress; after completion, listen to the audio, inspect both faces and framing, then download.

If the result is not what you expected

• Faces or roles look wrong: use one unobstructed, well-lit face per photo and check the left/right assignment. A new generation costs credits and may still vary.
• A performer is cropped: try 16:9 and a reference where both people stay visible. Vertical framing can leave less room for a wide duo scene.
• The soundtrack matches but lips do not: copying audio does not retime the generated mouth movements. Review before sharing; choose a different performance reference if timing matters.
• A task is still processing: check its existing entry in My Videos before submitting again. A network interruption alone does not mean it failed. Confirmed failed tasks return their charged credits.

When this workflow fits

Use this page when you want separate control over two identities, a performance reference and the final soundtrack. It generates a new performance rather than a frame-by-frame face replacement. For exact choreography, perfect lip-sync or a finished edit with captions, expect additional review or editing.

What is the Hotel Lobby trend?

The Hotel Lobby trend is a short-video format inspired by Quavo and Takeoff’s HOTEL LOBBY performance on A COLORS SHOW: two performers, an orange backdrop and a microphone hanging between them. In AI versions, people replace the original duo with themselves and a friend or another pair of characters. Hotel Lobby is the song’s name; the recognizable scene is a performance studio.

Hotel Lobby AI helps you make your version of the duo trend. This generator combines one photo per performer with a built-in or uploaded performance reference. Compare the real example above, then choose your portraits. Faces, movement and timing can vary; reference-audio mode preserves the source soundtrack, not exact lip-sync.

Choose your reference video and portraits

• Use a 4–15 second MP4, up to 30 MB, with both performers in view. A steady shot makes the scene and each person’s position easier to specify than a clip with frequent cuts.
• Upload one person per image: left performer first, right performer second. Choose sharp, well-lit faces without sunglasses or objects covering them. JPG, PNG and WebP are supported, up to 10 MB each.
• The built-in clip supplies the orange studio and hanging microphone reference. To change the performance, upload footage you have permission to adapt. Use portraits with permission from the people depicted.

Hotel Lobby AI prompt: what to add

The generator already tells the model which photo belongs to each performer and to follow your clip’s scene and movements. Use Additional direction for a specific detail, such as framing or expression:

Keep both performers in a medium-wide shot with the microphone between them. Use relaxed expressions and subtle hand gestures. Keep the orange background visible throughout.

How much does a Hotel Lobby AI video cost?

Required inputs

1 MP4 reference + 2 portraits

Video model

MiniMax H3

Aspect ratios

16:9 / 4:3 / 1:1 / 3:4 / 9:16 / 21:9

Audio

Keep reference audio or choose AI-generated sound.

Billing

Reference + output duration; varies by resolution

Output length

4 / 8 / 12 / 15 seconds

Output resolution

768P / 2K

¹ Credits cover both reference and output duration: round up (reference seconds + output seconds) × the resolution rate to a whole credit. Reference duration is rounded up to whole seconds first; the two portraits add no duration. Check the estimate above before generating.

This is a separate billing example, not the cost of the 4.5-second showcase video above.

The one-time 310-credit pack costs US$9.90. Starting with no credits, it covers 2 videos at 768P or 1 at 2K using the example settings above (not both). Remaining credits stay in your balance. This is a pack purchase, not a per-video cash price.