GPT Image 2AI 提示词
Bilingual AI Portrait Prompt Comparison — Sora 2 Prompt
Generate a split-image comparison: two nearly identical photorealistic editorial portraits of a businessman, proving AI consistency across languages with overlaid prompts.
提示词 · EN
type
对比图subject
通过不同语言的等效提示词渲染出的两次相同的写实编辑肖像layout
orientation
垂直堆叠sections
item 1
position
顶部count
1language
英语overlay_text
A Japanese businessman in a dark navy three-piece suit, sharp jawline, confident eyes, mid-30s, standing with arms crossed slightly off-center to the right, facing camera, modern glass-office at dawn with Tokyo skyline visible through windows, soft golden hour side lighting from upper left, warm color temperature, 85mm f/1.4 lens, shallow depth of field with creamy bokeh, three-quarter view, bust shot, photorealistic editorial portrait, cinematic color grading. Format 16:9.image
photorealistic portrait of {argument name="subject" default="a Japanese businessman"} in his mid-30s wearing a {argument name="suit color" default="dark navy"} three-piece suit with white shirt, dark tie, white pocket square, and wristwatch, sharp jawline, confident eyes, styled dark hair, standing with arms crossed, placed slightly right of center, facing the camera in a three-quarter bust shot inside a modern high-rise glass office at dawn, Tokyo skyline visible through floor-to-ceiling windows, desk with an open laptop and a small plant on the left, shelving and warm interior lights on the right, soft golden-hour side light from the upper left, warm color temperature, shallow depth of field, creamy bokeh, cinematic editorial realism, 85mm f/1.4 look, aspect ratio 16:9item 2
position
底部count
1language
日语overlay_text
浓绀のスリーピーススーツを着た日本人ビジネスマン、シャープな颚のラインと自信のある眼差し。30代半ば、両腕を胸の前で交差させ少し右寄りに立ち、カメラ正面を向いている。ガラス张りの现代的なオフィス、窓の外には夜明けの东京のスカイラインが见える。左上からの柔软なゴールデンアワーのサイド光、暖色系の色温度、85mm f/1.4レンズ、浅い被写界深度でクリーミーなボケ、3/4ビューのバストショット、フォトリアルなエディトリアルポートレート、シネマティックなカラーグレーディング、Format 16:9image
the same composition and styling as the top image, nearly identical photorealistic editorial portrait, preserving the businessman, office setting, dawn skyline, crossed arms, lighting, lens look, shallow depth of field, and cinematic grading to demonstrate minimal visual difference between English and Japanese promptsbackground
每个部分填充一半画布，无分割线，文本块位于每个面板的左上角count_total_sections
2style
简洁的社交媒体对比图，直接在生成的图像上方显示提示词文本intent
演示 GPT-image-2 通过英语和日语提示词能产生非常相似的结果关于这个提示词
Generate a split-image comparison: two nearly identical photorealistic editorial portraits of a businessman, proving AI consistency across languages with overlaid prompts. 它更适合作为 GPT Image 2 的概念艺术起点：先保留画面结构、主体关系和风格约束，再替换成你的品牌、人物或场景。
这条提示词使用结构化 JSON 组织信息，包含 type、subject、layout、style、intent 等字段。保留这种层级能让模型更清楚地区分画面主题、布局、界面元素和细节约束。
使用时建议先小幅修改主体、场景、镜头和色调，再生成多个版本对比构图与细节。这样页面内容对用户有实际帮助，也避免把模型名或标签机械堆在正文里。
如何使用这个提示词
- Copy the full prompt code including the layout structure and image generation arguments.
- Paste it into a capable AI image model like Sora 2 that supports parameterized prompts and overlaid text.
- Customize variables such as the subject, suit color, or camera angle by editing the
{argument}fields. - Run the prompt to produce a vertically stacked output with the English prompt on top and Japanese on the bottom.
- Tweak and regenerate until the portraits match closely and the text renders clearly.
