Skip to content

Examples

Watch our RainDance AI character demo

Watch our MiniMax H3 demonstration with two fictional AI characters. The silent 4:3 video is 27.04 seconds long and illustrates the style you can create with your own authorized portraits.

Choose your quality and submit your video in the generator. Your completed download is private.

Character references used in this demonstration

Fictional adult character reference image used for the dock role in the RainDance AI demonstration
Reference A — dock character
Fictional adult character reference image used for the waterfront performer role in the RainDance AI demonstration
Reference B — waterfront performer
Silent
Watch the silent RainDance AI character demonstration

Save the silent MP4. In TikTok, create a post and upload the file. Tap Add sound, select the Raindance sound available to your account, and adjust the music start to match the clip. Preview the timing before publishing. The opening second has been removed; music alignment is manual.

See how the character references shape the result

Two fictional full-body images provided the character references for this demonstration. One guides the dock character and the other the waterfront performer, including their appearance and clothing across the generated scenes.

Watch the frames between the best moments

The major pier, water and beach scenes remain recognizable. Some expressions, movements and internal cut times differ from the reference; this example is not an exact frame-by-frame reconstruction.

A generated interpretation of the coastal sequence

This demonstration uses reference material to guide the coastal setting, scene order and performance. The resulting frames are AI-generated, so expressions, movement, backgrounds and transitions may differ from the reference.

How this demonstration was made

This demonstration combines three MiniMax H3 reference-to-video outputs generated using the two fictional character images shown above. Reference A guides the dock character and reference B guides the waterfront performer. We trimmed the opening second, cropped the frame to 4:3 and prepared a silent version for display. The completed demonstration lasts 27.04 seconds; character details and movement may vary across AI-generated shots.