Hotel Lobby AI video examples

Two photos, two ways to create a performance. Compare a photo-and-prompt video with a video guided by the original movement and soundtrack. Both recorded tests below use the same fictional adult portraits.

Which scenes do these examples help you choose?

These examples demonstrate a shared studio performance: two people, a hanging microphone, a plain orange set and singing- or rap-style gestures. Use them to assess a friends' duo, a couple's playful performance or a short fictional-character stage concept. The portrait layout suits mobile feeds; the landscape layout gives the stage more space on a wider screen.

Hotel Lobby AI refers to this duo performance format. These clips do not demonstrate hotel advertising, accommodation walkthroughs or travel footage. The photo-only case creates a new performance; the reference-driven case uses a selected original-video excerpt for movement and audio. Neither is a guarantee of exact choreography or singing synchronization. Pet ideas and two versions of the same adult are experimental directions; the recorded source-to-output case on this page verifies only the shown pair of distinct fictional adult performers.

The same two input photos

Both cases use these AI-generated fictional adults. These are test inputs, not customer photos.

First fictional adult reference: curly-haired performer in a green sweater and jeans
Photo A: requested left performer.
Second fictional adult reference: brown-haired performer in a cream blouse and jeans
Photo B: requested right performer.

Example 1: two photos + prompt, without an original video

This case uses two AI-generated fictional adult portraits, submitted through the deployed Hotel Lobby AI application on October 1, 2026. The photos above are the inputs for both recorded cases, not photos of real customers. This first case uses MiniMax H3 with photos and creative direction; no original video or original-song recording was supplied. Its audio track is part of the generated output.

Generation conditions

The recorded request selected 5 seconds, 768P and portrait 9:16, and charged 50 internal test credits once. The delivered MP4 is 768 × 1344 pixels and 5.167 seconds long, with an audio track. These are measured output properties, not exact matches to the requested ratio or duration. This case is a single recorded test, not a success-rate measurement.

Compare the references with the output: both distinct performers appear together under the hanging microphone, with the green sweater and cream blouse carried into the orange scene. Inspect facial likeness and movement yourself. This case does not establish exact song performance, lyric synchronization or a guarantee that every run preserves identity equally well.

To make your own friend or couple clip, upload one clear authorized photo per adult, choose the orange template and portrait format, and describe a simple alternating performance. The current studio lets you select the model and shows the charge before submission. Read the scenario walkthrough or try photo-only generation.

Example 2: two photos + original movement and audio

This October 11, 2026 application test uses the same two fictional portraits shown above, plus a selected Hotel Lobby reference excerpt. Seedance 2.5 uses the original footage to guide gestures, camera framing and scene timing. The delivered video includes the corresponding original audio, rather than a newly invented soundtrack.

Enable sound to compare. Photo A remains the green-sweater performer on the left; Photo B is the cream-blouse performer on the right. This test used the 10–15 second window of the built-in reference, selected 5 seconds at 480p, and charged 400 internal test credits. The final MP4 measures 854 × 480 pixels and exactly 5 seconds, with a 5-second audio track.

The model returned slightly shorter footage; a small timing adjustment retained the returned actions and aligned them with the selected soundtrack. This is one successful test, not a measured overall success rate. Identity, choreography and lip sync may still vary. Original-audio generation is subject to provider approval; this example does not grant music usage rights.

Try the original-video reference workflow →

What is different between these two videos?

Generation conditionsExample 1: photo + promptExample 2: original-video reference
InputsTwo photos and creative directionThe same two photos, selected video excerpt and matching audio
MovementA newly generated performanceGuided by the original performance
AudioGenerated audio; no original recording suppliedThe selected original soundtrack
Measured output5.167 seconds · 768 × 13445 seconds · 854 × 480

More Hotel Lobby video examples: portrait and landscape

Friends on the mic

Portrait format · Two performers. One orange stage.

This short portrait clip runs for approximately five seconds, with a delivered frame size of 768 × 1344 pixels. The tall composition keeps both performers visible beneath the central microphone and gives you a useful reference for a mobile viewing layout. Watch how the subjects share the frame, then compare the wider example alongside it.

To prepare your own references, use two separate, clear photos with one authorized adult subject in each. Aim for similar lighting and avoid cropped or obscured faces. These are preparation tips, rather than a claim about the source photos used for this sample. The orange backdrop, hanging microphone and two-performer arrangement define the scene; your references and creative direction guide the characters, styling and movement. Individual results can vary, so review the completed clip before sharing.

The full studio moment

Landscape format · Enable sound in the player.

This landscape clip runs for approximately ten seconds and is delivered at 1600 × 900 pixels. It includes an audio track; unmute the player to listen. The wider frame provides more room around the performers, while the longer running time lets you inspect movement and continuity beyond a short opening moment. The dimensions describe this published file, rather than a verified record of its original generation settings.

For a similar layout in the studio, choose a wide frame and use your creative direction to describe how the two subjects should perform. The shared orange scene and central microphone keep the format recognizable, while clothing, gestures and performance details can change between requests. You do not need to submit an existing video: start with two reference photos you have permission to use. Compare both examples, choose the format that suits your intended presentation, and check the displayed credit charge before submitting your own request.

Questions about Hotel Lobby AI video examples

Which examples include source photos?

The two recorded cases above use the same fictional source portraits, shown above both cases. The additional gallery clips do not include source-photo pairs or a verified likeness comparison. For your own video, upload only photos you have permission to use, obtain each depicted adult’s consent and review the result before sharing.

Which format should I pick?

Choose a portrait frame for mobile short-form feeds, or a landscape frame for presentations and wider screens. Both use the same two-performer scene. Compare the examples above, then choose the available format, duration and resolution in the creator studio. The studio displays the credit charge before you submit.

Examples illustrate the format; individual generated results can vary. For your own inputs, follow the two-photo preparation checklist or the visual creation walkthrough.

Choose a reference video template

Keep the Hotel Lobby duo stage, or choose a specific performance and preview its original reference before uploading your photos.

Review each page's inputs, credit estimate and limitations. Motion and likeness vary; exact replication is not guaranteed.

Create your video →