Documento original
GPT Image 2 · zh
This is a soft, pastel-style anime illustration in a cute VTuber fan art style. The image depicts a shy, petite girl with short, light brown twin tails, fluffy bangs, a single ahoge (a cowlick), and small hair accessories including a heart-shaped hair clip. She has warm, large brown eyes, an extremely shy expression, flushed cheeks, and teardrop-shaped beads of sweat; her small mouth trembles slightly. She is sitting awkwardly on a small wooden stool, knees together, hands clenched tightly on her knees, clearly very uncomfortable with her posture. She wears an oversized cream-colored hoodie, a dark skirt, black socks, and black low-top sneakers with white laces. An arm with a dark blue sleeve extends from the left edge of the image, pointing at her with a teasing or accusatory tone. Comic-style stress symbols are added around the girl: a red, graffiti-like cloud of frustration above her head, two wavy, trembling marks near her body, and six blue sweat droplets floating around her. The bottom right corner features a white speech bubble containing the Japanese text "あうぅ…！姿态…悪い？" (Aww…! Is my…sitting posture really that bad?). The background is a softly lit pastel room. On the left is a cream-colored beanbag chair, and the bookshelves are filled with books, figurines, plush toys, and picture frames. On the right is a desk with a monitor, a PC, and a microphone stand. Behind the desk are wall art and character sketches. A pink blanket is draped over the furniture on the far right, and a light-colored rug covers the wooden floor. The room is warmly lit. The overall color scheme is soft pink and beige, with clean lines and delicate cel shadows. The characters are slightly chibi-style, creating a warm and slightly comical atmosphere.Cómo usar
- Copia el prompt original completo, incluidos los parámetros.
- Adapta el tema, el texto y la composición a tus necesidades.
- Usa una herramienta externa compatible; comprueba el acceso y el resultado.
Antes de empezar
- Comprueba si necesita una imagen de referencia. La compatibilidad no demuestra qué modelo generó el ejemplo.
- Revisa nombres, cifras y textos; el resultado puede variar entre versiones del modelo.
