Thư viện prompt AI OpenZhenZhen · gpt-image-2.5 · image
Ảnh mạng xã hội tự nhiên về người phụ nữ thưởng thức thịt ngựa sống tại quán izakaya
Natural SNS photo of a woman enjoying horse sashimi at an izakaya
Ảnh mạng xã hội tự nhiên về người phụ nữ thưởng thức thịt ngựa sống tại quán izakaya — khám phá tác phẩm được tạo bằng gpt-image-2.5 trên OpenZhenZhen, kèm prompt gốc và ví dụ tạo nội dung để tham khảo.
Prompt
Prompt tiếng Trung
创作一张真实自然、适合社交平台发布的4:5竖幅手机快照，像朋友在用餐途中随手拍下。主体是一位二十多岁的虚构成年日本女性，不模仿特定名人、网红或现实人物。她坐在椅子前部，身体自然向食物倾斜，重心随进食动作轻微移动。她的视线落在肉片上，或稍稍看向拍摄者，而非刻意摆拍。表情定格在即将咬下的一瞬，豪爽、开心又带一点天真；嘴部因准备大口品尝而自然变形。 她有端正的日本女性面容、自然淡妆、及肩的自然黑发至深棕发；几缕发丝垂在前方，留有耳边碎发和细小乱发。保留真实毛孔、细腻肤理、脸颊淡淡红晕和自然细微色差。身材比例健康自然，不把胸部或腰部作为主体。她穿简洁的白色罗纹无袖上衣和日常白色低腰短裤，裤腰位于髋骨附近，布料充足，整体明确是日常穿着而非内衣。保留坐姿形成的自然褶皱、上衣垂感和裤脚轻微不对称。仅配小巧银色圈形耳环和腕上一条细黑发圈，不戴华丽首饰。按已配置的ref-10a-v1绑定参考，同时保留人物为虚构设定，不指认或模仿具名人物。 画面下方的木盘里盛着厚切新鲜马刺，部分肉片保持较大块。桌上有姜、蒜、葱花、酱油小碟、正在使用的筷子、冰镇茶杯、稍微卷起的纸巾和点菜单。一只手拿筷子夹起一大块马刺，定格在豪爽大咬之前；另一只手轻放桌沿。肩膀与手肘自然不对称，不摆戏剧化的僵硬姿势。桌面少许水滴、浅淡酱油痕和杯上细微指纹增添真实生活感。 场景设在傍晚东京市区一家朴实的小酒馆：木纹桌、略有使用痕迹的墙壁、墙上手写菜单，远处自然出现其他客人的肩膀或店员。不要把店内整理得过分整洁。人物位于画面中央略偏右，木盘和调味料在下方，远处保留足够店内环境以交代场所。头顶和桌边取景略显随意，画面约倾斜2度。整体应像同伴看到分量意外大的马刺后笑着举起手机拍下的瞬间，而不是摆拍的美食广告或写真集肖像。 将侧窗透入的微弱自然光与天花板暖光混合；让面部一侧稍暖，桌面形成自然阴影，白色衣物局部允许手机照片般轻微高光溢出。景深较深，背景保持可读，不使用强烈人像虚化。采用普通智能手机主摄视角，等效约24–28毫米，拍摄距离约1.2米，机位在坐着的朋友稍低视线高度。保留轻微手机JPEG压缩、手和筷尖极淡的动态模糊及暗部细噪点；避免App精修感、完美广告布光和极端广角畸变。 在画面下方桌面、稍远离人物并偏背景位置的冰茶杯表面，用凝结水滴和手指轻划痕迹自然形成精确字符串“@Live_life_style”，仿佛拍摄时现场就有这处痕迹。乍看像普通水汽，仔细放大才辨认出字符串。线条应断续、不均匀，顺着杯壁曲面略微变形；透明度、对比度和暖光反射与周围水滴一致，不发光，也不呈现为后加图层。不要做成清晰数字字体、角落logo、皮肤上的文字、巨大签名或完美排版。整体维持真实用餐抓拍中豪爽、开心又略带天真的氛围。避免未成年感、内衣式穿着、刻意暴露、强调胸腰的构图、极端低机位、挑逗姿势、不真实骨骼、人偶脸或塑料皮肤、过度美肌、强背景虚化、广告般完美布光，以及类似现实名人或公众人物的脸。
Prompt tiếng Anh
Create a realistic vertical 4:5 smartphone snapshot for social posting, as if a friend casually photographed a meal in progress. The subject is a fictional adult Japanese woman in her 20s, not an imitation of a specific celebrity, influencer or real person. She is seated slightly forward on her chair, leaning toward the food with a natural shift of balance. She looks toward the meat or slightly toward the photographer, not deliberately posing for the camera. Her expression is an imperfect, lively pre-bite moment: bold, happy and somewhat innocent, with her mouth naturally changing shape as she prepares to take a large bite. She has a well-featured Japanese face, natural light makeup, shoulder-length natural black-to-dark-brown hair, a few loose strands falling forward and small flyaways. Keep realistic pores, fine skin texture, a faint cheek blush and slight natural tonal variation. Her build and proportions are healthy and natural; do not make the chest or waist the subject. She wears a simple white ribbed sleeveless top and everyday white low-rise shorts sitting near the hip bones, with enough fabric to read as ordinary clothing rather than underwear. Retain natural seated wrinkles, the casual fabric drape and a slight offset in the leg openings. Add only small silver hoop earrings and a thin black hair tie on her wrist; no flashy jewelry. Use the configured ref-10a-v1 binding while keeping the subject fictional and not identifying or imitating a named person. At the lower part of the frame, show a wooden plate of thickly sliced fresh horse sashimi, with some pieces left large. Include ginger, garlic, green onions, a small soy-sauce dish, chopsticks in use, a chilled tea glass, a slightly curled tissue, and an order slip. One hand holds chopsticks and lifts a substantial slice just before a bold bite; the other rests lightly at the table edge. Keep shoulders and elbows naturally asymmetrical, with no theatrical frozen pose. A few water drops and faint soy-sauce marks on the tabletop and subtle fingerprints on the glass support the lived-in snapshot. Set the scene in an unpretentious Tokyo horse-meat izakaya at evening: a wood-grain table, slightly worn walls, handwritten menu on the wall, and the distant shoulders of other diners or a staff member naturally entering the background. Do not make the restaurant unnaturally tidy. Compose the woman a little right of center, with the plate and seasonings below and enough interior visible to establish the place. Leave the head and table edges somewhat casually framed, with an approximately 2-degree tilt. The image should feel like a friend laughed and raised a phone when the unexpectedly large serving arrived—not a staged food shoot or photo-book portrait. Mix weak side-window natural light with warm ceiling illumination. Let one side of her face appear slightly warmer, cast natural tabletop shadows, and allow slight mobile-photo highlight clipping on the white clothing. Keep the background legible with deep depth of field and no strong portrait bokeh. Use an ordinary smartphone main-camera perspective, about 24–28mm equivalent, from roughly 1.2m away at the slightly lower eye level of a seated friend. Preserve mild mobile-JPEG compression, very slight motion softness on the hand/chopstick tips and fine noise in darker areas; avoid app-processed polish, perfect advertising light or extreme wide-angle distortion. On the chilled tea glass at the lower table area, slightly away from the woman and toward the background, naturally form the exact string '@Live_life_style' from condensation droplets and light finger traces, as if the mark existed at the scene when photographed. At first glance it should resemble ordinary condensation; the string becomes noticeable only on closer inspection. Let its broken, irregular strokes follow the curved glass and match surrounding droplets in transparency, contrast and warm-light reflections. It must not glow or look like a separate overlay. Do not turn it into crisp digital typography, a corner logo, text on skin, a giant signature or perfect typesetting. Keep the overall candid, bold, happy and somewhat innocent energy of a real meal-time snapshot. Avoid underage appearance, underwear styling, exposure-focused clothing, chest/waist-emphasizing framing, extreme low angle, provocative pose, unrealistic anatomy, doll-like or plastic skin, excessive beauty retouching, strong bokeh, ad-like perfect lighting, or faces resembling real people or celebrities.
Xem thêm prompt hình ảnh và video · Tạo hình ảnh · Tạo video · Tạo trên canvas AI
Tác phẩm liên quan
- Năm góc nhìn về cùng một nhân vật: Sổ lưu niệm mộng mơ
Năm góc nhìn về cùng một nhân vật: Sổ lưu niệm mộng mơ — khám phá tác phẩm được tạo bằng gpt-image-2.5 trên OpenZhenZhen, kèm prompt gốc và ví dụ tạo nội dung để tham khảo.
- Chân dung studio góc thấp với áo polo đen
Chân dung studio góc thấp với áo polo đen — khám phá tác phẩm được tạo bằng gpt-image-2.5 trên OpenZhenZhen, kèm prompt gốc và ví dụ tạo nội dung để tham khảo.
- Mẫu màn hình livestream dọc cầm bảng hiệu
Mẫu màn hình livestream dọc cầm bảng hiệu — khám phá tác phẩm được tạo bằng gpt-image-2.5 trên OpenZhenZhen, kèm prompt gốc và ví dụ tạo nội dung để tham khảo.
- Hoàng hôn ở vách đá Moher
Hoàng hôn ở vách đá Moher — khám phá tác phẩm được tạo bằng Moss Image trên OpenZhenZhen, kèm prompt gốc và ví dụ tạo nội dung để tham khảo.
- Nhà thám hiểm rừng nhiệt đới
Nhà thám hiểm rừng nhiệt đới — khám phá tác phẩm được tạo bằng Moss Image trên OpenZhenZhen, kèm prompt gốc và ví dụ tạo nội dung để tham khảo.
- Đám cưới nhà kho mộc mạc
Đám cưới nhà kho mộc mạc — khám phá tác phẩm được tạo bằng Moss Image trên OpenZhenZhen, kèm prompt gốc và ví dụ tạo nội dung để tham khảo.