OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

男子从社交帖子卡片中跃出

Man Leaping Out of a Social Post Card

创作超现实电影感画面：一位明确成年、虚构的男性从手机屏幕中的通用社交媒体帖子卡片向外迈出，人物与卡片内部肖像为同一ref-10a-v1身份。人物穿白色黑色印花卫衣、蓝色牛仔连帽外层及宽松牛仔裤，湿润凌乱的头发，神情平静俏皮；一条腿向观者伸出形成强烈但解剖可信的3D透视。手机屏幕在跨出位置呈受控裂纹、玻璃碎片与少量像素方块飞散，避免伤害或血迹。帖子卡片显示账号名Xaroonx、同一人物头像和简化互动图标，不复制Instagram标志或界面商标。自然肤质、电影感光线和景深，主体清晰，碎片不遮脸。

男子从社交帖子卡片中跃出 — AI image creation
男子从社交帖子卡片中跃出

提示词 / Prompt

中文提示词

以用户上传的参考图作为主要可见人物的面部身份依据。在不与身份参考冲突的情况下，保留来源的场景、造型、服装、姿势、构图、光线和视觉风格。必须实际绑定 identity_reference 文件 ref-10a-v1。创作电影感超现实合成画面：一位明确成年的虚构男性，穿白色简洁黑色图案卫衣、蓝色牛仔连帽外层和宽松牛仔长裤；头发湿润凌乱，神情平静而俏皮。他正从手机中的通用社交媒体帖子卡片向外迈出。帖子内肖像和小头像中的面部均须与 ref-10a-v1 身份一致。卡片采用原创、中性的社交帖子布局，准确显示账号名“Xaroonx”和少量简化互动图标；不得复刻 Instagram 标志、品牌界面或其完整商业外观。人物一条腿向观者伸出，形成强烈但解剖可信的透视；四肢数量正确，表情保持平静俏皮。屏幕边界处添加受控裂纹、少量半透明飞散玻璃片和克制的像素碎块，不出现受伤或血迹。采用自然肤质、电影景深与平衡的戏剧光线；面部和服装清晰，碎片不得遮挡脸部。避免其他人物、重复肢体、账号名之外难辨文字、标志和水印。

English prompt

Use the uploaded reference image as the facial identity reference for the main visible person. Keep the source scene, styling, clothing, pose, composition, lighting and visual style unless they conflict with the identity reference. Attach the required identity_reference file ref-10a-v1. Create a cinematic surreal composite of one clearly adult fictional man wearing a white sweatshirt with a simple black graphic, a blue denim hooded layer and loose denim pants; he has damp tousled hair and a calm, playful expression. He steps out of a smartphone's generic social-media post card. Show the same ref-10a-v1 face in the post portrait and small avatar. The card uses a neutral, original social-post layout, displays the exact handle “Xaroonx” and a few simple interaction icons; do not reproduce Instagram's logo, branded interface or exact trade dress. Reposition the man so one leg extends toward the viewer with strong but anatomically plausible perspective; preserve a believable number of limbs and a calm playful expression. Add controlled cracked glass, a few flying translucent shards and restrained pixel fragments at the screen boundary, without injury or blood. Use photoreal natural skin texture, cinematic depth of field and balanced dramatic lighting; keep face and clothing clear and prevent shards from crossing the face. Avoid extra people, duplicate limbs, unreadable text beyond the supplied handle, logos and watermark.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作