OpenZhenZhen AI Prompt Gallery · MiniMax H3 · video

Moonlit Window, Tea Shadows

Moonlit Window, Tea Shadows — explore this MiniMax H3 creation on OpenZhenZhen, with its original prompt and a reusable generation example.

Moonlit Window, Tea Shadows — AI video creation
Moonlit Window, Tea Shadows

Prompt

Chinese prompt

生成一段20秒、9:16竖屏的真人电影感画面,场景始终连贯,发生在夜间临水琴阁。唯一女性角色为成年女性,必须全程与提供的身份参考图一致:保持她的面容、肤色、黑色盘发、额前碎发和优雅身形。她穿象牙白古典长衣、深绛红长裙与披帛,服装始终完整且前后一致。参考图仅用于人物外貌。开场时,她在木窗旁的烛光中坐着弹古琴。她弹完最后一个音,稍作停顿,转向走近的第一人称镜头。她起身从镜头旁走过,深红披帛短暂掠过镜头;镜头让她先行,迟半拍再转身跟上。她继续走向窗边,转身离开时顺手关上半开的窗扇,窗外圆月、薄雾笼罩的水面和远处灯影短暂入画。她不停步,从附近茶案取起一盏备好的茶,向跟随而来的镜头递去,随后继续往里走。她穿过真实存在的轻纱帘,进入更温暖的内室,在矮榻边坐下。镜头跟入并自然停稳。短暂安静后,她抬眼看向镜头;轻纱从画面边缘缓缓飘入,逐渐遮住她并覆盖整个视野,以真实遮挡结束,不使用淡出。琴阁空间保持连续,采用第一人称镜头,并始终稍晚于她的动作作出反应。动作克制自然;不切镜、不环绕、不突然变焦、不摆姿势、不添加额外动作。低曝光、浅景深、柔和烛光、微弱冷色月光轮廓、真实阴影与丝绸质感。无字幕、无标志、无音乐。

English prompt

Create a 20-second, 9:16 live-action cinematic scene in one continuous nighttime waterside pavilion. The sole woman is an adult and must match the supplied identity reference image throughout: preserve her face, skin tone, black updo, loose fringe, and graceful build. Dress her in an ivory classical robe with a deep-crimson skirt and draped sash; keep her clothing intact and consistent. The reference image is for her appearance only. Begin with her seated at a guqin in warm candlelight beside a wooden window. She finishes playing, pauses, and turns toward the approaching first-person camera. She rises and walks past the camera, her crimson sash briefly brushing the lens; the camera lets her pass, then follows a beat later. She continues to the window and closes its partly open panel as she turns away, with a full moon, misty water, and distant lights briefly visible outside. Without stopping, she takes a prepared teacup from a nearby table and offers it back toward the following camera, then continues inward. She passes through a real gauze curtain into a warmer inner room and sits at the edge of a low couch. The camera follows and settles naturally. After a quiet pause she lifts her eyes toward the camera; gauze drifting in from the edge gradually obscures her and fills the view, ending with physical occlusion rather than a fade. Keep the pavilion spatially continuous and the camera in first person, always reacting slightly after her movements. Use restrained, natural motion; no cuts, orbiting, sudden zooms, posing, or extra gestures. Low exposure, shallow depth of field, soft candle glow, subtle cool moonlit edges, realistic shadows and silk texture. No captions, logos, or music.

Topics and styles

View API model

Explore more AI image and video prompts · Try image generation · Try video generation · Create on the AI canvas