01
One Beat — Multilingual Hip-Hop MV
A 24-second rooftop music video with fixed four-person casting from reference images and precise multilingual lip-sync.
Create with this directionAI VIDEO MODEL
Direct a longer story from words, images, clips, and audio references — then steer where it starts, how it moves, and where it lands in the video workspace.
This background video was created with Wan 3.0.
Create a videoCREATIVE DIRECTIONS
Each direction pairs a story idea with reference material and choices about motion, camera, and sound.
The videos below were created with Wan 3.0.
01
A 24-second rooftop music video with fixed four-person casting from reference images and precise multilingual lip-sync.
Create with this direction02
A 30-second motion-graphics chase built on two character references, handheld camera language, and a single consistent glowing ball.
Create with this direction03
An Eastern fantasy flight over a celestial festival in retro Chinese calendar style, from FPV dive to ribbon-climax finale.
Create with this direction04
A cold-humor Route 66 scene: a silent cowgirl fills her horse with fresh grass while the station worker’s worldview collapses.
Create with this directionMODEL GUIDE
Bring words, images, clips, and audio references into one creative brief, then shape the subject, motion, and sound until the direction is specific enough to revise shot by shot.
Wan 3.0 accepts up to 10 reference images, 5 video clips, and 5 audio tracks in one generation, and clips can run up to 30 seconds, so an idea keeps its structure from opening frame to final beat.
CAPABILITIES
PROMPT ANATOMY
A useful prompt gives each creative variable a job, then keeps the direction specific enough to revise.
WORKFLOWS
READY TO SEND
MORE DIRECTIONS
This lower wall curates Seedance 2.5 video inspiration for planning your Wan 3.0 scenes.