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

蓝门旁的深夜疲惫街头人像

Late-Night Weary Street Portrait by the Blue Door

9:16低光纪实街头人像：一位明确成年的20岁日本女性深夜坐在瓷砖入口处，背靠蓝色金属百叶门旁的墙，戴耳机、抱着白色花纹托特包。她穿宽松白T恤、黑色短裤、袜子和厚底靴，长黑发遮住部分眼睛；灰卷帘门与虚化红色路锥衬托安静城市夜景。人物身份默认参考标识为 ref-10a-v1。

蓝门旁的深夜疲惫街头人像 — AI image creation
蓝门旁的深夜疲惫街头人像

提示词 / Prompt

中文提示词

创作一张自然写实的9:16竖版低光街头纪实人像，主体是一位明确成年的虚构20岁日本女性。将她的面部身份绑定到 ref-10a-v1 固定人物身份参考图（默认值：ref-10a-v1 固定人物身份参考图）。她有柔和椭圆脸、精致小鼻子、自然粉色嘴唇和深棕大眼，长直略凌乱的黑发垂下并遮住部分眼睛；神情平静而疲惫，目光低垂。保留自然的不对称、可见毛孔和轻微皮肤瑕疵。她坐在浅色方砖地面上，背靠蓝色金属百叶门旁的墙，膝盖弯曲、双腿放松，躯干微微前倾，低头但面部仍部分可见。一只手放在膝上大号白色小黑纹托特包旁，另一只手自然垂在大腿附近。她戴黑色耳罩式耳机，穿宽松纯白棉T恤、黑色短裤、黑袜和厚底黑色系带靴。采用安静、自然且尊重主体的纪实抓拍感，不表现为偷拍或窥视镜头；服装呈现日常状态，不刻意强调身体。环境包括白色马赛克墙砖、灰色卷帘门、空荡的铺砖通道，以及下方前景中虚化的红色交通锥。使用昏暗人造街灯，呈现智能手机视频静帧般的真实质感和轻微低光颗粒；保持自然比例与真实皮肤，不做塑料感磨皮。不要文字或水印。

English prompt

Create a naturalistic vertical 9:16 low-light street-documentary portrait of one clearly adult, fictional 20-year-old Japanese woman. Bind her facial identity to ref-10a-v1 固定人物身份参考图 (default: ref-10a-v1 固定人物身份参考图). She has a soft oval face, delicate nose, natural pink lips, large dark-brown eyes partly veiled by long straight, slightly mussed black hair, a calm tired expression and a lowered gaze. Keep natural asymmetry, visible pores and slight skin imperfections. Place her seated on pale square tiles, back resting against the wall beside a large blue metal louvered door; knees bent, legs relaxed, torso leaning slightly forward, head lowered while her face remains partly visible. One hand rests near a large white tote with a small black pattern on her lap, the other relaxed near her thigh. She wears a loose plain white cotton T-shirt, black shorts, black socks, chunky black lace-up boots and black over-ear headphones. Frame as a quiet, candid but respectful documentary moment, not a covert or voyeuristic shot; keep the clothing natural and non-emphasized. Show white mosaic wall tiles, gray rolling shutter, empty tiled walkway and a softly blurred red traffic cone in the lower foreground. Dim artificial street lighting, realistic smartphone-video-still character and subtle low-light grain; retain natural proportions and realistic skin, no plastic retouching, text or watermark.

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