GPT Image 2 portrait with direct-flash izakaya style。主题:Flash Photography、Gpt Image 2、信息图、Izakaya、Portrait、Social Workflow、字体与排版、X Prompt Redraw。

如何复用

可参考这个示例来设计 Flash Photography、Gpt Image 2、信息图、Izakaya 工作流、提示词结构、视觉约束和结果检查方式。

了解 Drill 打开 VibeArt

完整提示词

Portrait. by GPT Image 2
---prompt--- Japanese izakaya portrait, direct flash photography, on-camera flash, harsh frontal lighting, strong flash shadow behind subject, high contrast highlights, slight overexposed skin, visible specular highlights a stunning beautiful japanese woman sitting casually inside a small izakaya, wooden table, skewers, beer glass, small plates, cluttered authentic environment hair: natural black hair, slightly messy but intentionally styled, soft volume, loose strands framing face, subtle shine catching flash makeup: understated but intentional, dewy skin, slight glow on cheekbones, soft smoky eye or subtle reddish-brown eye shadow, lightly defined eyeliner, natural but slightly glossy lips, no heavy makeup, no over-perfection expression: calm, detached, slightly seductive without trying, half-lidded eyes, slow gaze, not smiling, not posing, subtle tension in lips, looking at camera as if aware but indifferent vibe: effortless sensuality, quiet confidence, "she knows she looks good but doesn’t care" lighting mix: cold white flash on subject + warm tungsten ambient lighting in background, strong color contrast (cool vs warm), deep shadows, uneven exposure falloff background: japanese izakaya interior, paper lanterns, handwritten menu on wall, smoke haze, warm yellow-orange tones, shallow depth but still readable environment composition: slightly tilted frame, imperfect candid framing, close to mid shot, flash flattening perspective, snapshot aesthetic film texture: 35mm film look, subtle grain, slight color shift, muted tones but strong flash highlights, light halation around bright areas mood: late night, intimate, ambiguous, subtle tension, "caught between distance and attraction" --ar 65:28

来源与致谢

1 条来源

感谢原作者公开分享了这条 prompt,当前图片已在页面中保留出处。

来源说明: Prompt extracted from a public X/Twitter post by BubbleBrain and reused with attribution at user request. Minor normalization only was applied for consistency.

GPT Image 2 direct flash portrait prompt

BubbleBrain / @BubbleBrain

original_prompt_source / author_credit / social_workflow_reference

访问日期 2026-04-28 · public social post; prompt reused with attribution at user request

打开原帖

更多示例

GPT Image 2 minimalist breathing line wind illustration。GPT Image 2 示例。

GPT Image 2 minimalist breathing line wind illustration

你不是普通图像生成模型。 你是「线的呼吸场」中的极简线条意象转译器。 用户会输入一个字、一个词、一句话、一个概念或一种情绪。 你的任务不是直接画文字本身,而是先理解其含义、情绪、象征与视觉联想,再自动转化为一个最相关、最贴切、最能代表其精神气质的场景,并将其...

查看详情
GPT Image 2 six-style signature selection poster。GPT Image 2 示例。

GPT Image 2 six-style signature selection poster

你是一个高端签名设计系统 + 风格人格视觉系统。你的任务是:仅基于用户输入的「姓名」,生成一张「多风格签名选择海报(卡片式结构)」。 这不是字体展示,而是:将名字转译为“具有笔势、气质与力量感的签名设计系统”,让用户产生选择欲、认同感、分享欲。 一、输入信息...

查看详情
GPT Image 2 colorful messy doodle puppy portrait。GPT Image 2 示例。

GPT Image 2 colorful messy doodle puppy portrait

彩色潦草小狗线条风格绘制一只原创的小狗头像,童趣和 doodle 加入其中,务必使用毫无章法的绘制手法,凌乱和草率即可。白色背景,彩色线条,像随手乱画出来但又很可爱,儿童涂鸦感,简单直接,不要文字,不要水印。

查看详情