OpenZhenZhen AI Prompt Gallery · gpt-image-2.5 · image

Mirror selfie in a sunlit gym

Mirror selfie in a sunlit gym — explore this gpt-image-2.5 creation on OpenZhenZhen, with its original prompt and a reusable generation example.

Mirror selfie in a sunlit gym — AI image creation
Mirror selfie in a sunlit gym

Prompt

Chinese prompt

将 identity_reference ref-10a-v1 的完整照片作为视觉输入,创作一张写实、自然抓拍感的健身房镜前自拍照片,采用纵向9:16画幅。保持同一女性可辨识的身份、面部结构、肤色和五官特征,不刻意改动。她站在健身房自由重量区的大镜子前,穿黑色长袖拉链连体衣和浅色运动鞋;头发松散地盘成发髻,妆容精致自然。她一只手拿水瓶,另一只手拿智能手机;手机上清楚可见泛化的健身应用界面,但不要出现个人姓名、健康数据或编造的运动指标。镜中反射需清楚呈现健身房细节,包括力量架和一排哑铃。自然日光从健身房大窗户照入,画面干净真实。采用纪录片式写实生活方式摄影,氛围轻松随性,人体比例自然,镜面与反射关系可信。采用从头到脚的站立构图,让完整服装和浅色运动鞋都清楚可见;同时确保面部、手持物品和健身房环境在纵向画幅内清楚可辨。不要添加其他人物、无关文字、标志或水印。

English prompt

Use the full identity_reference ref-10a-v1 photograph as the visual input. Create a realistic, candid vertical 9:16 gym mirror-selfie photograph of the same woman. Preserve her recognizable identity, facial structure, skin tone and facial features without intentional changes. Show her standing in the free-weights area in front of a large mirror, wearing a black long-sleeved zipped jumpsuit and light-colored sneakers. Style her hair in a loose bun and give her polished, natural-looking makeup. She holds a water bottle in one hand and a smartphone in the other; make a generic workout app interface visibly present on the phone, with no personal name, health data or fabricated performance metrics. In the mirror reflection, clearly show gym details including a power rack and a rack of dumbbells. Let natural daylight stream through large gym windows for a clean, authentic look. Use documentary-style realistic lifestyle photography, a relaxed candid feel, natural proportions and believable mirror/reflection geometry. Frame her standing head-to-toe so the complete outfit and light-colored sneakers are visible; keep her face, both held objects and the gym context legible within the vertical frame. No additional people, unrelated text, logos or watermarks.

View API model

Explore more AI image and video prompts · Try image generation · Try video generation · Create on the AI canvas