提示词/GPT Image 提示词/提示词详情图片提示词九宫格 AI 美颜脸拼贴生成一张 3×3 拼贴，包含相似精致东亚奇幻自拍肖像，用于研究或戏仿重复的 AI 基础脸美学。此提示词使用了参考图片立即使用点赞7浏览2,906收藏6保存或分享此提示词，下次可以轻松找到分享提示词提示词复制Create a square 3×3 collage of exactly 9 close-up AI-generated female portrait selfies, all with the same highly polished GPT-image aesthetic: youthful East Asian fantasy/cosplay beauty, large glossy eyes, porcelain skin, soft blush, small nose, slightly parted lips, and an uncanny near-identical face structure across all panels. Canvas is a 1:1 square, divided into 9 equal tiles with no gutters or labels. Each tile is a tight head-and-shoulders crop, mostly frontal or slight three-quarter angle, cinematic shallow depth of field, warm sunlight and soft bokeh backgrounds, ultra-detailed skin and hair, realistic yet subtly doll-like. Include exactly 9 distinct variations: 1) black-haired girl with straight bangs, open-mouth smile, warm indoor festival lighting; 2) elegant traditional fantasy girl with dark updo, red floral forehead mark, red dangling earrings, pink makeup; 3) brown-haired girl wearing a red headband, fingers near cheek, backlit outdoors; 4) short dark-haired girl with red headband, soft smile, pale sunlit background; 5) blue-haired anime-cosplay girl with bangs, glittery skin texture, parted lips; 6) purple-haired girl with bangs, star earrings, hand touching cheek, gentle smile; 7) black-haired girl with pink/red headband and star hair accessory, warm cosplay look; 8) dark brown-haired girl with wispy bangs and star earrings, soft neutral expression; 9) brown-haired girl with loose strands and red headband, serious delicate gaze. Emphasize repeated base-model facial similarity, glossy retouched realism, social-media portrait collage composition, high detail, vibrant but natural colors. No text, no watermark, no borders beyond the tile divisions.展开提示词
