OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

奔跑主体与原创3D迷你版本的城市奇幻场景

Urban fantasy scene with a running figure and an original 3D mini version

A vertical 9:20 photoreal-to-stylized composite in which the adult reference subject sprints through a colorful city street while extending one hand toward a tiny original 3D toy-like double.

奔跑主体与原创3D迷你版本的城市奇幻场景 — AI image creation
奔跑主体与原创3D迷你版本的城市奇幻场景

提示词 / Prompt

中文提示词

创作一张9:20竖幅构图的画面：参考图中的成年主体在色彩鲜明、风格化的城市街道中全速奔跑，并保留参考图中的人物身份特征。主体身体前倾，表情既专注又愉快，忠实穿着参考图人物的现实服装，同时去除可见标志和文字；主体伸出一只手，朝向一个微小、原创、玩具感的3D迷你分身。迷你分身带着滑稽惊慌的表情，迈着细小的动画步伐在风格化路面上向前奔跑；明确表现为原创CGI角色，不复制任何特许动画风格，也不将其塑造成第二个真实人物身份。环境为色彩丰富的原创3D动画城市街道，包含圆润建筑、俏皮车辆、愉快的背景行人和细致的风格化路面。通过匹配透视、接触阴影、色彩和运动方向，让写实成年主体与风格化迷你角色自然融合。成年主体保持真实的解剖结构、手部和面部特征以及自然动作。无标志或水印。

English prompt

Create a vertical 9:20 composition featuring the adult subject from the reference image sprinting at full speed through a colorful, stylized city street, with the subject's identity preserved from the reference. The subject leans forward with an intense but joyful expression, wears the reference subject's real-world outfit faithfully adapted without visible logos or text, and extends one hand toward a tiny, original, toy-like 3D miniature double. The miniature has a comically panicked expression and tiny animated legs as it runs ahead on stylized pavement; keep it clearly an original CGI character, not a copied franchise style or a second real-person identity. Build the environment as a colorful original 3D-animated city street with rounded buildings, playful cars, cheerful background pedestrians, and detailed stylized pavement. Blend the photoreal adult subject and stylized miniature convincingly through matching perspective, contact shadows, color, and motion direction. Preserve realistic anatomy, hands, facial features, and natural motion in the adult subject. No logos or watermarks.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作