GPT Image 2AI 提示词
Natural Light Over-the-Shoulder Portrait — Sora 2 Prompt
Create a vibrant urban portrait featuring a young woman's playful over-the-shoulder glance, enhanced by golden natural light and soft lens flare.

提示词 · EN
Medium close-up shot, eye-level camera angle, focusing on a young woman in a seven-eighths body frame. She is seated with a slight lean, her upper body twisted to the right rear while her head turns toward the camera, forming a classic over-the-shoulder glance. Her gaze is direct, clear, and slightly playful. She has fluffy brown shoulder-length hair with natural bangs, light natural makeup with subtle eyeliner, and nude-pink lips. Natural daylight filter with golden light coming from the upper left, creating soft backlighting contours on her face and body, with warm highlights on the right side and natural shadow transitions on the left for enhanced dimensionality. Bright natural light with slight lens flare creating an afternoon sunlight atmosphere. Eye-level shooting angle, the subject positioned slightly right of center, with zebra crossing and road lines forming natural leading lines toward the subject. Background is a city street with roads, crosswalks, greenery, and distant vehicles, moderately blurred but recognizable. She wears a military green camouflage hoodie, black shorts, white high-top sneakers with white mid-calf socks, and a black canvas backpack with orange decorative stripes. Overall style is street casual, youthful, energetic urban aesthetic.
关于这个提示词
Create a vibrant urban portrait featuring a young woman's playful over-the-shoulder glance, enhanced by golden natural light and soft lens flare. 它更适合作为 GPT Image 2 的概念艺术起点:先保留画面结构、主体关系和风格约束,再替换成你的品牌、人物或场景。
常见用途包括 portraits、advanced、portrait、street photography。这些标签只作为创作方向参考,实际生成时仍应以提示词结构和预览图为准。
使用时建议先小幅修改主体、场景、镜头和色调,再生成多个版本对比构图与细节。这样页面内容对用户有实际帮助,也避免把模型名或标签机械堆在正文里。
如何使用这个提示词
- Copy the prompt text provided.
- Paste it into your preferred AI image or video generation tool (e.g., Sora, Midjourney, Stable Diffusion).
- Customize details like clothing, background, or lighting to match your vision.
- Run the generation and iterate as needed to achieve the perfect result.



