该提示词用于即梦图片4.5模型生成多人合照。它要求将3张参考图中的人物合成到一张图片中，采用纯白简洁背景、紧密依偎的高角度俯拍自拍构图，头部比例夸张放大、符合日韩视觉自拍风格，光线柔和、色调清新，细节清晰呈现发丝与衣物褶皱，氛围温馨治愈。

中文版提示词

即梦图片4.5模型提示词：将3张参考图中的人物生成在一张图片中，纯白色简洁背景，3个人紧密依偎靠在一起自拍，高角度俯拍视角，每个人都直视镜头，表情自然生动，头部比例夸张放大，符合日韩视觉自拍风格；光线柔和均匀、无硬阴影，画面色调清新柔和，细节清晰呈现发丝纹理和衣服轻褶皱，整体氛围温馨治愈。

英文版提示词

Jimeng Image 4.5 model prompt: Generate the people from 3 reference images into a single image, with a pure white minimal background, the 3 people snuggling closely together in a selfie, shot from a high angle looking down, each person looking directly at the camera with natural, lively expressions, exaggeratedly enlarged head proportions in a Japanese-Korean visual selfie style; soft, even lighting without harsh shadows, a fresh and gentle color tone, details clearly showing hair texture and light clothing wrinkles, with an overall warm and healing atmosphere.

🛠️ **适用 AI 工具**：即梦、Midjourney、Stable Diffusion、DALL·E、豆包、Nano Banana