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

雷电将军误入魔兽世界的副本冒险分镜

Raiden Shogun's Misadventure into a Warcraft Dungeon Storyboard

9:16竖版3 x 6分镜：雷电将军意外来到魔兽世界，记录她组队、探索并挑战副本的一天。

雷电将军误入魔兽世界的副本冒险分镜 — AI image creation
雷电将军误入魔兽世界的副本冒险分镜

提示词 / Prompt

中文提示词

创作一张9:16竖版3 x 6分镜，讲述《原神》的雷电将军意外进入《魔兽世界》，度过一天刷副本/团队副本的交叉世界冒险。用18格清晰连续的画面，从左到右、从上到下阅读；角色身份、服装和视觉连续性保持一致。描绘成年虚构角色雷电将军，面部身份匹配配置参考ref-10a-v1，同时保留她可辨认的经典角色造型与身份。 分镜依次呈现来到奇幻世界、结识队伍、准备开团、进入副本、穿越险境、与敌群战斗、和队友配合、抵达首领、激烈首领战、取得胜利以及一天结束时的安静片刻。每格内容有区分又共同组成同一天的故事；背景采用可辨认的《魔兽世界》风格奇幻环境、副本建筑、怪物与团队活动，并融入雷电将军的雷元素特征。采用富有表现力的游戏同人插画，轮廓清楚、动作易读、法术与战斗效果灵动，分格间距统一，色彩丰富但克制，细节完成度高。清楚保持3列6行网格，输出一张完整图像而非多张独立图片。若加入字幕或文字，须少量且清晰；不要虚构界面截图或堆砌杂乱元素。避免角色外观不一致、重复肢体、乱码文字、水印或无关场景。

English prompt

Create a 9:16 vertical 3 x 6 storyboard about Raiden Shogun from Genshin Impact unexpectedly entering the world of World of Warcraft and spending a day farming/raiding dungeons. Show one continuous crossover adventure across 18 clear panels, read left-to-right and top-to-bottom, with consistent character identity, outfit and visual continuity. Depict the adult fictional Raiden Shogun matching configured facial identity ref-10a-v1 while preserving her recognizable canonical character styling and role. Let the panels progress through arrival in the fantasy world, meeting a party, preparing for a raid, entering a dungeon, navigating hazards, fighting groups of enemies, coordinating with allies, reaching a boss, an intense boss encounter, victory and a quiet end-of-day moment. Keep each panel visually distinct but connected as one day's story; use recognizable World of Warcraft-inspired fantasy environments, dungeon architecture, creatures and raid-party activity alongside Raiden Shogun's electro-themed presence. Use expressive game-fan-art illustration with clear silhouettes, readable action, dynamic spell and combat effects, coherent panel gutters, rich but controlled color and polished detail. Preserve a clear 3-column by 6-row grid, one single image rather than separate outputs. Keep captions and lettering minimal and legible if used; do not invent interface screens or clutter. Avoid inconsistent character appearance, duplicated limbs, unreadable text, watermarks or unrelated scenes.

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