Galerie de prompts IA OpenZhenZhen · gpt-image-2.5 · image

Invites de portraits hyperréalistes mignons style K-drama

Hyper-realistic Cute K-Drama Style Portrait Prompts

Invites de portraits hyperréalistes mignons style K-drama — découvrez cette création avec gpt-image-2.5 sur OpenZhenZhen, son prompt original et un exemple de génération à réutiliser.

Invites de portraits hyperréalistes mignons style K-drama — AI image creation
Invites de portraits hyperréalistes mignons style K-drama

Prompt

Prompt chinois

使用配置的ref-10a-v1作为两个镜头中同一成年女性的身份参考。保持可辨认的面部比例与特征;将100%理解为严格身份保持要求,而非数值化保证。创作一组浪漫韩剧风超写实双镜头肖像。镜头1:她甜美地看向镜头,双手在胸前摆出清晰爱心,穿黑色、酒红色和白色格纹西装外套,内搭黑色上衣并佩戴精致银项链;高马尾系飘动的黑白缎带。镜头2:同一女性温柔微笑,双手做韩式手指比心,穿黑白粗花呢夹克和深红色上衣;高马尾系浅蓝与白色长缎带。两镜头的手势都要完整清晰。背景为奢华温暖室内,采用金色环境光、奶油般电影感动态虚化背景、优雅景深和细小漂浮金色光粒。呈现韩国美妆社论与韩剧宣传照品质,保留HDR、f/1.4人像镜头质感、面部清晰对焦、浅景深、高级调色、真实皮肤和自然发丝。避免身份漂移、过度磨皮、塑料皮肤、解剖或手部变形、多余手指、重复肢体、面部裁切、过饱和、强硬光线、不自然妆容、文字、标志、水印、噪点和伪影。

Prompt anglais

Use configured ref-10a-v1 as the identity reference for the same adult woman in both shots. Preserve recognizable facial proportions and features; interpret 100% as a strict identity-preservation instruction, not a numerical guarantee. Create an ultra-realistic, romantic K-drama-inspired two-shot portrait series. Shot 1: she smiles sweetly toward camera and forms a clear heart with both hands in front of her chest, wearing a black, burgundy, and white plaid blazer over a black top with a delicate silver necklace; her high ponytail has flowing black-and-white satin ribbons. Shot 2: the same woman smiles softly and makes Korean finger-heart gestures with both hands, wearing a black-and-white tweed jacket over a deep-red top; her high ponytail has long pastel-blue and white satin ribbons. Keep both hand gestures visible and well formed. Use a luxurious warm indoor setting, golden ambient light, creamy cinematic motion-blur background, elegant depth, and tiny floating gold light particles. Aim for Korean beauty-editorial/K-drama promotional quality with HDR, f/1.4 portrait-lens character, sharp facial focus, shallow depth, premium grading, realistic skin, and natural hair strands. Avoid identity drift, over-retouching, plastic skin, anatomy or hand distortion, extra fingers, duplicate limbs, cropped face, oversaturation, harsh light, unnatural makeup, text, logos, watermarks, noise, and artifacts.

Voir le modèle API

Découvrir plus de prompts image et vidéo · Générer des images · Générer des vidéos · Créer sur le canevas IA