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

武士刀切西瓜与菠萝双场景动作广告

Samurai Sword Cutting Watermelon and Pineapple Dual-Scene Action Ad

一张双画幅高端运动商业摄影作品，同一位明确25岁以上、身份绑定固定参考的成年男子分别完成两种空中切水果动作：黑色运动装场景中右手持刀切开悬浮西瓜，黄色运动卫衣与黑裤场景中右手持刀切开悬浮菠萝。深黑影棚、体积雾、电影灯光、金属刀刃反光、果汁与水滴的定格飞溅及潮湿反光地面共同强化动态能量。全身动作清晰，水果内部细节锐利；无伤害、血腥或人体接触。

武士刀切西瓜与菠萝双场景动作广告 — AI image creation
武士刀切西瓜与菠萝双场景动作广告

提示词 / Prompt

中文提示词

创作一张高端双画幅动作美食广告。两幅画面中均为同一位明确年满25岁的成年健壮男子，并绑定ref-10a-v1、保持相同身份。第一幅：他穿合身黑色圆领运动卫衣和慢跑裤，右手在空中挥刀切开悬浮的新鲜西瓜，清楚呈现鲜红果肉与籽。第二幅：他穿亮黄色圆领卫衣和黑色长裤，右手在空中挥刀切开悬浮的菠萝，清楚呈现金黄色果肉与果芯。表现受控且富有戏剧性的武士刀切水果动作，刀刃只能接触水果，绝不能碰到人物。加入受控的皇冠形果汁与水花、悬浮水滴和少量干净果块，不出现受伤或血液。场景设于深黑影棚，配有体积雾、潮湿反光地面和武士刀金属高光。两个画幅均采用高反差电影灯光与黑色奢华商业广告氛围；水果细节保持锐利，只让选定水滴带有动态模糊。人物均完整全身入镜，水果切面清晰可辨。保持真实人体结构，并确保模特与刀刃之间留有安全距离。不要血液、受伤、额外人物、标志、水印或未经支持的分辨率/渲染引擎宣称。

English prompt

Create a premium two-panel action-food campaign featuring one clearly adult athletic man, age 25 or older, identity-bound to ref-10a-v1 with the same identity in both panels. Panel one: he wears a fitted black crewneck athletic sweatshirt and joggers, and uses his right hand to make a mid-air cut through a suspended fresh watermelon, with vivid red flesh and seeds visible. Panel two: he wears a bright yellow crewneck athletic sweatshirt and black pants, and uses his right hand to make a mid-air cut through a suspended pineapple, with golden flesh and core visible. Show a controlled, theatrical katana demonstration cutting fruit only; the blade must never contact a person. Add controlled crown-shaped fruit-juice and water splashes, suspended droplets and a few clean fruit pieces, with no injury or blood. Set both panels in a deep-black studio with volumetric mist, a reflective wet floor and metallic katana highlights. Light both panels with dramatic high-contrast cinematic light and a black luxury-commercial mood; keep fruit detail sharp and motion blur limited to selected droplets. Keep each figure full-body and each fruit interior sharply readable. Maintain realistic anatomy and safe distance between model and blade. No blood, injury, extra people, logos, watermark or unsupported resolution/render-engine claims.

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