20

电影级 AI 模型对比网格
点击放大

示例图片

海报营销YouMindgpt-image-2posters-typography海报与排版

电影级 AI 模型对比网格

生成一张四宫格电影级对比图,展示不同 AI 模型对“俯瞰军队的史诗奇幻统治者”这一场景的输出效果。

分类
海报营销
模型
GPT Image 2
来源作者
V
原始语言
en
浏览量0
来源 ID
28064
发布时间
2026年7月8日
Studio
App 同款来源
完整提示词
目标:创建一个 2x2 的电影级对比网格,展示由四个图像模型渲染的同一史诗奇幻场景,每个象限均使用优雅的白色斜体衬线字体进行标注。

画布:宽屏 16:9 图像,尺寸约为 1200x675,精确划分为 4 个相等的面板,采用简洁的 2x2 布局,除面板间的自然缝隙外无边框。

布局:四个面板排列如下:左上角标注“Midjourney v8.2”,右上角标注“Seedream 5.0 Pro”,左下角标注“Muse Image”,右下角标注“GPT Image 2”。将每个标签放置在各自面板的左上角附近,使用小号白色斜体衬线字体。

主体与场景:在每个面板中,展示相同的构图概念:一位孤独的统治者或指挥官背对观众,站在高处的石制阳台或露台上,在一座宏伟的浅色石制堡垒城市中,向下方集结的庞大军队举手致意。人物身穿长款深灰色或柔和橄榄色斗篷/披风,内搭盔甲,可见后脑勺,背对观众面向军队。建筑宏伟、冷峻且古老:灰白色石墙、高塔、拱门、柱廊、庭院和仪式旗帜。军队由数千名排列整齐的士兵组成,庭院中可见部分骑兵或旗手。氛围庄重、宏大且具有电影感,如同奇幻战争电影。

面板细节:精确包含 4 个面板。面板 1,“Midjourney v8.2”:阳台上身披斗篷的统治者,非常对称的宏伟庭院,堡垒外远处的山脉,成排的士兵和旗帜,冷色调的去饱和蓝灰色光影。面板 2,“Seedream 5.0 Pro”:更近的阳台视角,右侧有拱形柱廊,密集的士兵人群填满庭院,朦胧的浅色石质氛围,指挥官举起右手。面板 3,“Muse Image”:左下视角,指挥官形象稍大且色调更深,举起戴手套的手,高大的矩形石制护栏,下方是网格状排列的庞大军队,堡垒墙壁在阴天背景下向远处延伸。面板 4,“GPT Image 2”:更朦胧柔和的视角,指挥官居中,斗篷背部有刺绣,举起的手部分呈半透明或有动态模糊,军队呈长条汇聚排列,前景有白色石柱和旗帜。

视觉风格:照片级真实感的电影剧照,高奇幻史诗尺度,去饱和冷色调分级,雾气弥漫的空气透视,柔和的阴天光线,浅色石质纹理,细腻的胶片颗粒,严肃的戏剧性基调,无现代物体。

场景的可自定义核心提示词:描绘 {argument name="main character" default="一位身披斗篷的盔甲统治者"} 俯瞰 {argument name="army description" default="一支数千人组成的庞大仪式化队列军队"},身处 {argument name="location" default="一座宏伟的白石奇幻堡垒城市"},采用 {argument name="lighting" default="冷色调去饱和阴天电影光影"},并标注对比 {argument name="model labels" default="Midjourney v8.2, Seedream 5.0 Pro, Muse Image, GPT Image 2"}。

约束:在所有四个面板中保持相同的场景类型,保留精确的 2x2 对比布局和 4 个标签,避免额外的说明文字,避免 Logo,避免水印,避免明亮饱和的色彩,并保持图像的电影感而非插画感。
多语言2

电影级 AI 模型对比网格

zh-CN

目标:创建一个 2x2 的电影级对比网格,展示由四个图像模型渲染的同一史诗奇幻场景,每个象限均使用优雅的白色斜体衬线字体进行标注。 画布:宽屏 16:9 图像,尺寸约为 1200x675,精确划分为 4 个相等的面板,采用简洁的 2x2 布局,除面板间的自然缝隙外无边框。 布局:四个面板排列如下:左上角标注“Midjourney v8.2”,右上角标注“Seedream 5.0 Pro”,左下角标注“Muse Image”,右下角标注“GPT Image 2”。将每个标签放置在各自面板的左上角附近,使用小号白色斜体衬线字体。 主体与场景:在每个面板中,展示相同的构图概念:一位孤独的统治者或指挥官背对观众,站在高处的石制阳台或露台上,在一座宏伟的浅色石制堡垒城市中,向下方集结的庞大军队举手致意。人物身穿长款深灰色或柔和橄榄色斗篷/披风,内搭盔甲,可见后脑勺,背对观众面向军队。建筑宏伟、冷峻且古老:灰白色石墙、高塔、拱门、柱廊、庭院和仪式旗帜。军队由数千名排列整齐的士兵组成,庭院中可见部分骑兵或旗手。氛围庄重、宏大且具有电影感,如同奇幻战争电影。 面板细节:精确包含 4 个面板。面板 1,“Midjourney v8.2”:阳台上身披斗篷的统治者,非常对称的宏伟庭院,堡垒外远处的山脉,成排的士兵和旗帜,冷色调的去饱和蓝灰色光影。面板 2,“Seedream 5.0 Pro”:更近的阳台视角,右侧有拱形柱廊,密集的士兵人群填满庭院,朦胧的浅色石质氛围,指挥官举起右手。面板 3,“Muse Image”:左下视角,指挥官形象稍大且色调更深,举起戴手套的手,高大的矩形石制护栏,下方是网格状排列的庞大军队,堡垒墙壁在阴天背景下向远处延伸。面板 4,“GPT Image 2”:更朦胧柔和的视角,指挥官居中,斗篷背部有刺绣,举起的手部分呈半透明或有动态模糊,军队呈长条汇聚排列,前景有白色石柱和旗帜。 视觉风格:照片级真实感的电影剧照,高奇幻史诗尺度,去饱和冷色调分级,雾气弥漫的空气透视,柔和的阴天光线,浅色石质纹理,细腻的胶片颗粒,严肃的戏剧性基调,无现代物体。 场景的可自定义核心提示词:描绘 {argument name="main character" default="一位身披斗篷的盔甲统治者"} 俯瞰 {argument name="army description" default="一支数千人组成的庞大仪式化队列军队"},身处 {argument name="location" default="一座宏伟的白石奇幻堡垒城市"},采用 {argument name="lighting" default="冷色调去饱和阴天电影光影"},并标注对比 {argument name="model labels" default="Midjourney v8.2, Seedream 5.0 Pro, Muse Image, GPT Image 2"}。 约束:在所有四个面板中保持相同的场景类型,保留精确的 2x2 对比布局和 4 个标签,避免额外的说明文字,避免 Logo,避免水印,避免明亮饱和的色彩,并保持图像的电影感而非插画感。

电影级 AI 模型对比网格

en

Goal: Create a cinematic 2x2 comparison grid showing the same dark medieval-fantasy scene rendered as if by four different image models, with a white italic label in the upper-left of each panel. Canvas: Wide 16:9 collage, four equal rectangular panels arranged in a 2 by 2 grid, no borders or gutters, each panel cropped like a film still. Cold desaturated blue-gray color grading, overcast sky, misty highland mountains, rocky cliffs, wet grass, distant valleys, dramatic natural light. Layout: Exactly 4 panels. Top-left label: “Midjourney v8.2”. Top-right label: “Seedream 5.0 Pro”. Bottom-left label: “Muse Image”. Bottom-right label: “GPT Image 2”. Use elegant white italic serif text, small-to-medium size, placed near the top-left corner of each panel with slight softness, no text boxes. Subject: Each panel depicts the same core scene: a rugged middle-aged male commander, {argument name="character type" default="a bearded medieval warlord"}, standing on a rocky mountain path with soldiers following behind him. He wears dark navy and black armor with steel plates, leather belts, medals, layered tunic or skirt armor, gloves, boots, and heavy weathered fantasy-war costume details. He carries {argument name="weapon" default="a long halberd or poleaxe"}. His expression is stern and battle-worn, with windblown hair and a serious posture. Panel details: Top-left panel should feel like the most polished cinematic fantasy still: close three-quarter view of the commander walking along a cliffside path, thick fur mantle around his shoulders, blue armor, sword or pole weapon partly visible, a column of soldiers blurred behind him, steep cliff rising to the right and foggy sea or valley to the left. Top-right panel should show a wider heroic full-body view: the commander stands on jagged rocks holding the halberd horizontally, distant troop formation marching on a mountain road at the right, vast gray valley and storm clouds behind him. Bottom-left panel should be a high-angle mountain pass scene: commander in the foreground looking downward, armored and holding a double-headed poleaxe, long visible line of troops snaking across the rocky landscape below, low clouds and mist over the mountains. Bottom-right panel should be moody and spacious: commander in dark armor standing on a rocky ledge with poleaxe angled downward, smaller soldiers behind him, towering cliff wall on the left, bleak valley on the right, heavy storm clouds. Visual style: Ultra-cinematic live-action fantasy realism, epic historical drama atmosphere, high detail armor and fabric textures, subdued contrast, shallow depth of field in some panels, realistic mist, no bright colors. The overall prompt theme is {argument name="scene theme" default="a grim medieval commander leading an army through a stormy mountain pass"}. Constraints: Keep exactly 4 labeled panels and exactly the four visible labels listed above. Avoid logos, watermarks, extra captions, modern objects, sci-fi elements, cartoon styling, or additional panels.