GPT Image 2.5 Sunburst 提示词 - 柔和窗光下的坐姿女仆角色扮演
- 画面比例
- auto
- 画质
- 1K
提示词
A natural, photorealistic indoor cosplay portrait in a horizontal 3:2 frame. An adult woman, 25 or older, sits sideways on a white oval shag rug, positioned on the right side of the composition. She has long straight black hair, dense blunt bangs, long face-framing strands, and a black-and-white ruffled maid headband with black cat-ear-shaped accents. Fair skin, blue-gray eyes, defined black eyeliner, long lashes, pink blush, and muted peach-pink lips. Her head is slightly tilted as she looks directly into the camera with a quiet, relaxed expression.
She wears a black maid-inspired costume with puffed short sleeves, a pointed white collar, white cuffs and buttons, a white apron with ruffled shoulder straps and hem, broad white wrist cuffs, and sheer smoky-gray thigh-high stockings with pale lace tops. Her legs are bent in a natural side-seated pose, one lower leg extending toward frame left and the other folded in front of her. One palm rests on the rug at frame right to support her weight; the other hand lightly holds a white shoulder ruffle. Preserve believable anatomy, natural weight distribution, and fabric creases.
Behind her is a cream loveseat with rounded nailhead-trimmed arms and white feet, backed by luminous white sheer curtains. A white drawer cabinet appears at frame right, while a small white table, stool, and gray bags occupy frame left. A pair of black patent pumps sits separately on the floor at the lower left. Retain the partially cropped circular light at the upper left and a soft blurred foreground obstruction along the far-left edge. Photograph from slightly above her seated face with a gentle downward angle, an estimated 35–50mm full-frame equivalent lens, and moderate depth of field resembling f/4–f/5.6. Soft window daylight, pale reflected fill, muted colors, gentle contrast, and visible hair, skin, fabric, and rug texture. She has just settled onto the floor; the camera catches her looking up while adjusting her costume.
使用这条图片提示词
- 01
从提示词开始
复制提示词,或点击生成按钮在生成器中打开。
- 02
加入你的想法
调整主体、场景或细节,按需添加参考素材。
- 03
生成前确认
检查模型与输出设置,每次生成的结果可能有所不同。







































