Original document
GPT Image 2 · en
A bright, anime-style close-up portrait depicts a girl standing outdoors in the rain, half-length, holding a transparent plastic umbrella. She has {argument name="hair color" default="Dark blue-black"} Her hair was tied in a low side ponytail with soft wisps and straight bangs. She wore delicate round-framed glasses and had bright, large, purplish-blue eyes. Her expression was calm, with a hint of amazement, as she gazed up at the leaves and raindrops. She wore a soaked white short-sleeved school uniform shirt with a dark blue plaid bow at the collar. The scene was filled with realistic water and light: fine rain fell, raindrops clung to the umbrella ribs, water droplets clung to her hair and shirt, and crystal-clear water droplets hung from the large green leaves in the upper right foreground. In the lower right and foreground, blooming white garden flowers dotted the scene, their dark, glossy leaves covered in raindrops. The use of cool blue and silver tones, complemented by soft pale purple highlights, and a background of diffused sunlight after or during the rain, created a strong sense of transparency, reflection, refraction, shimmering bokeh, and the glossy sheen of wet surfaces. The background should be an out-of-focus city garden or residential walkway, with fences and buildings blurred in the rain, while the girl herself remained clear. The film emphasizes cinematic composition, layered foreground leaves, ultra-fine raindrops, realistic translucent umbrella material, subtle contour lighting on wet hair, and an overall serene and breathtaking atmosphere reminiscent of a summer rain scene.How to use
- Copy the complete original prompt, including any parameters.
- Replace the subject, text and composition with your own requirements.
- Use a compatible external tool; check its current access conditions and review the result.
Before you start
- Check whether a reference image is required. A compatible model label does not mean this example was generated with that model.
- Keep names, figures and visible text accurate; results can vary between model versions.
