小动物视角天台手机追逐时尚肖像

来源：OpenNana / 作者：@MrLarus / 模型：ChatGPT / 类型：image

小动物视角天台手机追逐时尚肖像，适合短视频封面、AI生图、AI视频分镜和内容创作参考。模型：ChatGPT。

去 AI 视频生成工具提取/制作短视频封面

示例图

图片来自公开来源，仅作提示词效果参考；正式商用请按来源授权处理。

提示词正文

Create a realistic 9:16 ultra-low-angle fashion-action portrait from the POV of a tiny animal that has just stolen a young woman’s phone.

The camera is almost touching the rooftop ground, using a 16–20mm ultra-wide lens with strong foreshortening. A 20–23-year-old adult East Asian woman lunges toward the lens, reaching out as if trying to grab the phone back. Her hand, legs, and burgundy strappy heels appear dramatically enlarged in the foreground, while her face remains sharp and recognizable.

She has vivid wine-red hair tied in a high ponytail, with loose strands and flying hair showing fast movement. She wears a fitted black spaghetti-strap ruched mini dress with realistic fabric tension and natural upper-body support. Her expression is focused, annoyed, and slightly desperate, as if saying, “Give it back!” She looks directly at the camera with lively eyes, closed lips, and slightly lowered brows.

Bright blue sky, hard natural sunlight, rooftop concrete edges and minimal railings. Photorealistic skin texture, realistic anatomy, natural body proportions, believable balance and weight, candid split-second energy rather than a posed fashion shoot.

Create each image separately, not as a collage. Keep the same woman, face, hairstyle, outfit, shoes, lighting, and rooftop environment, while varying only the reaching gesture, leg position, distance from the lens, and stage of the chase.

短视频创作用法

• 做封面：保留主体、光线、构图、色彩关键词，改成自己的账号主题。
• 做AI视频：把静态画面拆成 3-5 个镜头，补充运镜、动作和转场。
• 做图文/口播：用提示词生成视觉钩子，再配合标题和脚本提高点击率。