Wan 3.0 AI動画ジェネレーター

Vertical 9:16, one continuous 15-second take, shallow depth of field, subtle handheld UGC feel. Medium close-up at eye level: woman late 20s, cream knit sweater, soft window light from camera-left. She speaks to camera with natural hand gestures — single shot, no cuts, no on-screen text. Native audio: clear voice forward, quiet home-office room tone, no music. Stable face throughout the full 15 seconds.

From the uploaded document, create one 20-second 16:9 product teaser. Focus on three beats only: problem, solution, proof — skip boilerplate and legal footer text. Clean modern SaaS ad look, subtle UI motion, no fake logos, no unreadable tiny on-screen text. One continuous flow, not slide-by-slide hard cuts. Optional native audio: calm professional voiceover tone, light ambient bed, no loud music.

Using Image1 for the woman's face, hair, and olive green bomber jacket, and Video1 for camera path, pacing, and handheld feel: Vertical 9:16, one continuous 8-second take. Match Video1 motion exactly — do not copy Video1 background or clothing. Image1 identity stable throughout. No on-screen text.

16:9 cinematic transition from the closed jar start frame to the open jar end frame. One smooth 8-second continuous move: lid release between frames — same jar silhouette and frost glass texture, cream revealed inside. No new packaging graphics, no on-screen text. Premium beauty lighting, photoreal materials.