视频 图片
登录
动画青葱丰收女主角

动画青葱丰收女主角

A detailed text prompt for generating a dynamic anime fantasy heroine proudly holding green onions in a bright harvest field.

观看135
点赞9
评论0
分享0
收藏0
引用0
翻译
Creating a vibrant anime-style full-body vertical illustration of one young female fantasy heroine standing in a lush green onion field under a brilliant summer sky. The character is the main subject, viewed from a low angle, centered slightly right, with dynamic wind blowing her very long {argument name="hair color" default="black hair with violet-purple gradients"} dramatically to the left. Her face is intentionally covered by a soft-edged square blur, while the rest of the image remains crisp and highly detailed. She wears an elegant white blouse with wide sleeves, a teal-green sleeveless outer robe/cape, a dark teal pleated skirt, a layered white overskirt, a navy waist sash, thigh-high white stockings, and ornate botanical accessories. Include exactly 8 visible outfit/accessory elements: 1 white blouse, 1 teal outer robe, 1 dark teal pleated skirt, 1 layered white overskirt, 1 navy waist sash, 1 floral chest brooch, 1 tassel-and-ring belt ornament, and 1 pair of white thigh-high stockings. Her right arm is raised high, proudly holding exactly 1 large bunch of fresh {argument name="vegetable" default="green onions"} with long green leaves and white stalks; her left hand hangs relaxed near her side. The setting is a refreshing harvest scene with neat rows of green onions filling the foreground and midground, distant blue-green mountains on both sides, and a vivid cobalt-blue sky with fluffy white clouds. Add scattered windblown green onion leaves and small leaf fragments across the sky for motion and freshness. Use high-detail Japanese anime illustration style, luminous clean colors, crisp linework, painterly highlights, flowing fabric, sparkling sunlight, and a heroic yet refreshing harvest atmosphere. Aspect ratio 3:4, no text, no watermark, no extra characters.

机器翻译 — 实际生成作品的是原文提示词。

✨ 生成 原始帖子 在 Hitboard 打开

如何使用这个提示词

复制提示词粘贴到 GPT Image 2。构图、光线与风格词请按原文保留,这是产出该结果的关键;换掉主体或场景细节让它成为你的,比例保持不变,若提示词提到参考图就一并附上。多生成两三次,最好的版本通常不是第一张。