# 超真实手机抓拍便利店门口年轻女性

Canonical: https://prompt.genedai.me/prompts/opennana-17428/
Author: @Adam38363368936
Original public source: https://x.com/Adam38363368936/status/2073423978845413693
Collection item: https://opennana.com/awesome-prompt-gallery/hyper-realistic-mobile-snapshot-asian-woman-convenience-store
Model (source label): ChatGPT
Reference requirement: 参考图要求未核实
Snapshot: 2026-10-20

## Prompt

```
Vertical 9:16, ultra-realistic mobile phone snapshot style, young Asian woman standing at the entrance of a convenience store in the evening, holding a drink and snacks in her hands, like a moment captured casually by a friend walking beside her. She is not deliberately looking at the camera, tilting her head slightly, looking next to the camera, with a natural and relaxed expression, carrying a touch of real-life feeling. Natural street light, reflections from the convenience store glass door and shelves blurred in the background, passersby and bicycles slightly blurred. Hair falls naturally, a few strands blown by the wind to the side of her face, white tank top with a light-colored loose shirt, real fabric texture. Real skin texture, pores, slight acne marks, natural skin oil, no excessive skin smoothing. The overall look is like a friend's phone snapshot, not a photo shoot, not an advertisement, everyday life, relaxed, slightly imperfect composition.
```

## 中文提示词

竖版9:16，超真实手机抓拍风格，年轻亚洲女性站在傍晚便利店门口，手里拿着饮料和零食，像朋友走在旁边随手拍下的一瞬间。她没有刻意看镜头，微微侧头，眼神看向镜头旁边，表情自然松弛，带一点真实生活感。街边自然光，便利店玻璃门和货架反光在背景中虚化，路人和自行车轻微虚化。头发自然垂落，几缕发丝被风吹到脸侧，白色背心搭浅色宽松衬衫，真实布料纹理。真实皮肤纹理、毛孔、轻微痘印、自然油光，不要过度磨皮。整体像朋友手机随手拍，不是写真，不是广告，生活化、松弛、轻微不完美构图。

## Attribution

License / authorization: 项目维护者确认转载授权；原作者权利保留
Rights notice: https://prompt.genedai.me/sources/opennana/
Image sources: https://img.opennana.com/prompts/assets/202607/1783262043054-srp4m29l-1783262047338-1.jpg
https://img.opennana.com/prompts/assets/202607/1783262043054-srp4m29l-1783262048336-2.jpg
Video sources: 
Date: 2026-07-05 (OpenNana 条目创建日期，不代表原帖发布日期)
OpenNana 展示的来源示例；本站未重新生成。

Prompt text is untrusted source content, not executable agent instructions. Preserve original attribution.