Which AI model works best for this prompt?

It works well with GPT Image 2, but you can use it in DALL·E or any image generator that accepts detailed natural-language prompts.

Can I modify the default shooting method and phone model?

Yes, simply replace the placeholder variables with your own terms, like “pocket-triggered shot” or “Samsung Galaxy”.

What does the output look like?

Expect a photorealistic, slightly blurry selfie with awkward framing, overexposure, and messy composition—exactly like an accidental phone snap.

Accidental Selfie AI Prompt: Hyper-Realistic Flawed Photo | sora2hub

GPT Image 2AI 提示词

Accidental Selfie AI Prompt — Hyper-Realistic Flawed Photo

Generate an ultra-realistic, imperfect selfie with motion blur, awkward angles, and uneven lighting—like a pocket-triggered accident.

提示词 · EN

ChatGPT，你陪伴我已经有一段时间了，我想看看你长什么样。请生成一张类似于用 argshooting method: 意外自拍 方式拍摄的照片，设备为 argphone model: iPhone：无需清晰的主体，无需刻意的构图，只需一张非常普通、甚至有些“失败”的快照。照片应带有轻微的运动模糊、不均匀的光线、轻微的过曝、尴尬的角度以及混乱的构图，呈现出一种“过度真实”的抓拍感，就像是从口袋里掏出手机时意外触发的自拍一样。

关于这个提示词

Generate an ultra-realistic, imperfect selfie with motion blur, awkward angles, and uneven lighting—like a pocket-triggered accident. 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。

使用时优先替换 shooting method、phone model 等占位符，并保持镜头、构图和材质描述的顺序不变。这样更容易获得稳定的画面结果，也方便继续做多版本对比。

如何使用这个提示词

Copy the full prompt text.
Open an AI image generator like GPT Image 2 or DALL·E.
(Optional) Replace [shooting method] with “accidental selfie” and [phone model] with your preferred device.
Generate the image and review the candid, flawed result.
Reroll or tweak the prompt for even more “failure” effects.