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

原创破招剑术高光动作分镜

Original Sword-Breaking Technique Highlight Action Storyboard

一张16:9横版3×4十二格动作分镜，展示原创青衫剑客在云雾山崖剑坪上逐招破势并最终定场。

原创破招剑术高光动作分镜 — AI image creation
原创破招剑术高光动作分镜

提示词 / Prompt

中文提示词

制作一张电影级写实武术动作分镜，单幅16:9横版画面，清晰排列为3×4的12格动作分镜，每格只使用少量原创中文标题（若能清晰呈现）。主体是一位原创且明确成年的青衫剑客，清瘦灵活、眼神锐利专注，穿青灰长衫，长发半束，手持朴素长剑。所有格子位于有山风的石质剑坪，旁边是风化旧剑庐，背景包含云雾山崖、石壁、竹林、落叶和尘土。动作按看破、切入、破势、连破、定场推进。12格依次表现：孤剑临阵；观察对手动势；切入剑招空隙；卸开刀势重量；沿枪路切入；截断棍势；截断长鞭环形轨迹；化解掌击；听风避箭；剑步连续破招；多道剑路归为一条决定性线路；最后定场收势。每格必须清楚表现不同的破招方式、距离、时机和兵器互动，不要只画火花碰撞或混乱群战。最后一格清楚看到对手在安全距离外倒地、兵器散落、胜负已定，但不要血腥。保持边框和格序整洁连贯，便于比较。排除具体系列、复制角色、标志、水印、过量文字、血腥和不可读杂乱信息。

English prompt

Create a cinematic realistic martial-arts action storyboard in a single 16:9 horizontal image arranged as a clear 3x4 grid of 12 cells, with minimal original Chinese title labels only if legible. Feature an original green-robed swordsman, clearly adult, lean and agile, with sharp focused eyes, a green-gray long robe, half-tied long hair, and a simple long sword. Set every cell on a windswept stone sword platform beside a weathered hut in a misty mountain landscape with cliffs, stone walls, bamboo, clouds, fallen leaves, and dust. Organize the action as discern, cut in, break posture, continuous breaking, and finale. The 12 distinct shots are: solitary sword facing an opponent; observing momentum; cutting an opening in a sword attack; unloading a saber's weight; entering the line of a spear; intercepting a staff; snapping a whip's circular motion; dispersing a palm strike; listening to the wind before an arrow; continuous sword footwork; many sword trajectories resolving into one decisive line; and a final grounded stance. Make each cell visibly demonstrate a different counter-technique, distance, timing, and weapon interaction rather than generic sparks or chaotic group fighting. In the last cell show defeated opponents at a safe distance, scattered weapons, and a clearly decided outcome without gore. Keep borders and cell order clean, coherent, and easy to compare. No named franchise, copied character, logos, watermark, excessive text, gore, or illegible clutter.

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