OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

博斯普鲁斯夜景餐厅中的绯红时尚肖像

Crimson Fashion Portrait at a Bosphorus Night Restaurant

一位绑定成年女性身份参考的虚构成年女性，坐在可眺望博斯普鲁斯海峡夜景的高端餐厅中，以暖光呈现精致晚间穿搭. A reusable image recipe made with gpt-image-2.5; explore the prompt and related works in the gallery.

博斯普鲁斯夜景餐厅中的绯红时尚肖像 — AI image creation
博斯普鲁斯夜景餐厅中的绯红时尚肖像

提示词 / Prompt

中文提示词

创作一张竖幅4:5的精致晚间生活方式时尚肖像，主角是一位明确成年的虚构女性，并始终绑定已配置的成年女性身份参考图ref-10a-v1。保留该配置参考所代表的身份，不要根据未提供的来源照片臆造面部、头发或皮肤细节。她舒适地坐在雅致餐桌旁，双腿自然交叠，肩部略转向镜头，带着温和自信的微笑并直视镜头。她穿剪裁合体的绯红色短款晚宴连衣裙，版型稳妥、面料不透明，搭配带图案的黑色薄丝袜和抛光酒红色及膝皮靴，鞋跟纤细但适度；仅佩戴简约耳钉，不加其他首饰。整体表达优雅且非性化，不刻意强调腿部或身体。场景为伊斯坦布尔一家可眺望夜间博斯普鲁斯海峡的高端餐厅：大窗户呈现亮灯的海峡大桥、深色水面及倒映其上的城市灯光；白色桌布餐桌与柔焦餐厅内部衬托主体。使用温暖室内环境光，在面部形成柔和悦目的光晕，皮肤纹理自然真实，并呈现高端手机生活方式编辑摄影气质。保持清楚的坐姿四分之三身构图。不要让额外人物成为焦点、可读招牌、标志、水印、美化改变，也不要无依据地承诺相机性能。

English prompt

Create a polished vertical 4:5 evening lifestyle-fashion portrait featuring one clearly adult fictional woman, consistently bound to configured identity reference ref-10a-v1. Preserve the identity represented by that configured reference without inventing facial, hair or skin details from an unavailable source photo. Show her seated comfortably at a refined dinner table, legs crossed naturally and shoulders turned slightly toward camera, with a gentle confident smile and direct gaze. Dress her in a tailored crimson red short cocktail dress with a secure, opaque fit, patterned sheer black tights and polished burgundy knee-high leather boots with a modest slim heel; add only minimal stud earrings, with no other jewelry. Keep the fashion presentation elegant and non-sexualized; do not emphasize the legs or body. Set the scene in a high-end Istanbul restaurant overlooking the Bosphorus at night: a large window reveals the illuminated bridge and dark water with city lights reflected on the surface; a white-clothed table and softly blurred restaurant interior frame the subject. Use warm ambient interior light with a gentle flattering glow on the face, realistic natural skin texture and a premium smartphone-lifestyle editorial feel. Maintain a clear three-quarter seated composition. No extra people in focus, readable signage, logos, watermark, beauty alteration, or unsupported camera-performance claims.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作