PROMPT RADAR / IMAGE平面设计

个人资料 / 头像 - Kawaii Academy Acrylic Sign Portrait

Generates a playful studio portrait of a blue-haired woman in kawaii fashion holding a colorful Japanese logo acrylic sign.

@splash_GL 0 热度 0 浏览
个人资料 / 头像 - Kawaii Academy Acrylic Sign PortraitORIGINAL VISUAL / 来源见文末
编辑先看

它为什么值得进入灵感库

先别急着照抄整段 Prompt:真正有用的是主体、空间和镜头之间的约束关系。 先规定信息层级和文字安全区,再谈视觉风格;无法落版的漂亮画面不是完整设计。

完整视觉 Prompt

Create a high-key vertical studio portrait photo of an adult Japanese woman standing against a plain soft light-gray background, framed from head to upper thighs. She has a short wavy bob haircut with black roots and vivid electric-blue dyed ends, straight bangs, pale blue translucent square glasses with exactly 2 lenses, and a gentle closed-mouth smile while tilting her head slightly. She wears a colorful one-piece swimsuit covered in cute kawaii sticker-style motifs such as bunnies, stars, strawberries, clouds, hearts, flowers, and pastel cartoon shapes. She holds a clear rectangular acrylic sign horizontally in front of her torso with both hands visible at the left and right edges; the sign has exactly 4 small round metal fasteners, one near each corner. On the sign, place a bold pop-idol Japanese logo with thick white and hot-pink lettering, black shadow outlines, sparkles, bubbles, lightning bolts, hearts, and stars; the main text should read {argument name="logo text" default="りぉみ 魔法学校"} with a small English subtitle underneath reading {argument name="subtitle text" default="RioMi Magic Academy"}. Use glossy magazine-quality photography, soft even lighting, realistic skin texture, shallow depth of field, crisp focus on the face and sign, playful Harajuku/kawaii color palette, no watermark, no extra people, no clutter.
查看原作者内容

不要一次改完所有变量

用 GPT Image 2 复刻时,先固定主体和构图,再逐次替换光线、色调或材质中的一个变量,这样更容易判断哪项描述真正起作用。

  1. 主体:Generates a playful studio portrait of a blue-haired woman in kawaii fashion holding a colorful Japanese logo acrylic sign.
  2. 视觉方向:社区实测摄影提示词;按原提示词的光线描述执行
  3. 镜头与构图:按原提示词的镜头与景别执行;按原提示词的构图与画幅执行
  4. 交付约束:auto,遵循原提示词中的避免项;不要水印、乱码、低清、畸形手、错误透视或主体漂移。
负向约束

遵循原提示词中的避免项;不要水印、乱码、低清、畸形手、错误透视或主体漂移。

模型建议

GPT Image 2 / Grok Imagine / Midjourney / Flux。先用原始 Prompt 建立基准结果,再根据模型对自然语言、镜头运动或文字排版的能力做局部改写。

来源与编辑原则

保留作者,也保留判断。

本页收录公开案例并提供中文索引、结构化拆解和复刻建议,不冒充原作者作品,也不替代原帖上下文。

作者:SPLASH GIRLS@AI Image · 收录于 2026年8月10日

访问原始来源

沿着这条线继续看