风中的地铁站台，一位美丽的女性身穿都市风格装扮。

Windy subway platform with a beautiful woman wearing a urban outfit. 1. Nano Banana Pro with reference 2. Grok Imagine 1.0 via API 3. GPT Image 2 via API with reference CR : Threads @hodumeru Prompt condensed by ChatGPT : Use the attached facial reference as the primary identity reference for the adult woman. Preserve identity precisely: eye/nose/mouth size and shape, eye spacing, forehead and face proportions, cheek/cheekbone contours, jawline, chin length, skin tone, natural asymmetry, and overall character. No stereotypical “AI beauty”: do not enlarge/sharpen eyes, shrink/narrow the face, or overly slim the chin. Keep head/neck at a normal angle. Keep the face undistorted and recognizable despite wind/framing. SCENE: Real city subway platform, late evening, cool bright station lighting. Tiled floor, columns, signage, ads, fixtures, distant commuters; lived-in, not futuristic or staged. She stands safely behind the platform line as a bright yellow subway train speeds past immediately behind her. Its body, windows, and lights remain discernible through strong natural horizontal motion blur. HAIR: Long straight hair below the waist, C-curled ends, long bangs. Train-induced wind sends hair dramatically in one direction, with strands at varied heights and some crossing her face; main strands remain distinct. Jacket and outfit hem flutter subtly. Wind feels like a real train wake, not a studio fan. OUTFIT: Rebellious street fashion: dark-gray fitted crop top under an open oversized black bomber; low-rise black cargo miniskirt with small pockets, belt, silver chain and metal details; black knee-high socks; chunky street boots; small black crossbody bag; silver necklace and subtle silver accents. MAKEUP: Korean “innocent pink” style, clear and natural. Authentic skin texture, rosy lips, no excessive retouching/artificial sheen; enhance without changing identity. PHYSIQUE: Full, mature adult proportions with natural womanly curves and built-in outfit padding. Natural upper-body volume, relatively slender waist, full pelvis/hips and visible waist-to-hip transition; skirt follows the pelvis. Maintain realistic skeletal structure, gravity, and anatomical balance; no exaggerated hourglass. POSE: Natural, not exaggerated. Weight shifted onto one leg while subtly bracing against the train gust. Relaxed shoulders/neck, calm chic expression, natural gaze at camera. Body naturally upright, not contorted. CAMERA/COMPOSITION: Vertical portrait, head to just above knees; subject fills frame with minimal empty space. Tilted camera creates a dynamic diagonal; platform and train cross frame at an angle. Close-range perspective without wide-angle distortion: no oversized head, receding body, protruding facial center, or narrowed chin. Natural head-to-body proportions. FOCUS/MOTION: Tack-sharp face and body, especially reference-accurate eyes, nose, mouth, and jawline. Hair strands frozen clearly while the yellow train behind has strong horizontal motion blur. STYLE: Ultra-realistic street-fashion photograph capturing a spontaneous moment as a real train passes, not a staged studio shoot. Natural skin/anatomy, realistic lighting, no beauty filters or excessive retouching. #aiart #aigenerated

风中的地铁站台，一位美丽的女性身穿都市风格装扮。 1. Nano Banana Pro 带参考图 2. Grok Imagine 1.0 通过 API 3. GPT Image 2 通过 API 带参考图 出处致谢：Threads @hodumeru 提示词由 ChatGPT 压缩： 以附带的面部参考图作为该成年女性的主要身份参照。精确保留身份特征：眼/鼻/嘴的大小与形状、眼距、额头与面部比例、脸颊/颧骨轮廓、下颌线、下巴长度、肤色、自然不对称性以及整体气质。不要刻板的“AI 美女”：不要放大/锐化眼睛、不要收缩/拉窄脸型、不要过度削尖下巴。头部/颈部保持正常角度。尽管有风与构图，脸部仍须不变形且可识别。 场景：真实的城市地铁站台，深夜，站台灯光清冷明亮。瓷砖地面、立柱、标识牌、广告、设施、远处的通勤者；生活化、非未来感、非摆拍。她安全地站在站台黄线后方，一列亮黄色的地铁列车紧贴她身后高速驶过。强烈的自然水平运动模糊下，车身、车窗与车灯仍应清晰可辨。 头发：过腰的黑色长发，发尾呈 C 卷，长刘海。列车带来的风将头发猛地吹向同一方向，发丝高度各异，部分掠过面部；主要发束保持清晰可辨。夹克与衣摆轻微飘动。风感像真实的列车尾流，而非影棚风扇。 穿搭：叛逆街头风：深灰色紧身短款上衣，外搭敞开的黑色超大飞行员夹克；低腰黑色工装超短裙，带小口袋、腰带、银色链条与金属细节；黑色及膝袜；厚底街头靴；小型黑色斜挎包；银色项链与低调银色配饰。 妆容：韩系“纯欲粉”风格，清透自然。真实皮肤纹理，唇色红润，无过度修图/人工光泽；在不改变身份的前提下进行美化。 体型：丰满成熟的成年比例，自然的女性曲线与服装自带的垫感。上身体量自然，腰部相对纤细，骨盆/臀部丰满且腰臀过渡明显；裙子顺应骨盆线条。保持真实的骨骼结构、重力感与解剖平衡；不要夸张的沙漏型。 姿势：自然，不夸张。重心偏向一侧腿，微微迎风而立。肩颈放松，表情冷静时髦，自然地望向镜头。身体自然直立，不扭曲。 镜头/构图：竖版人像，从头部到膝盖略上；主体填满画面，留白极少。倾斜的镜头形成动感对角线；站台与列车斜穿过画面。保持近距离视角且无广角畸变：不要出现头部过大、身体后缩、面部中心前突或下巴收窄的现象。头身比例自然。 对焦/动态：面部与身体极致清晰，尤其是与参考图一致的眼睛、鼻子、嘴巴与下颌线。发丝被清晰定格，而背景的黄色列车带有强烈的水平运动模糊。 风格：超写实街头摄影，捕捉一列真实列车驶过时的瞬间，而非影棚摆拍。自然肤质与解剖结构，真实光影，无美颜滤镜与过度修图。 #aiart #aigenerated