GPT Image 2AI 提示词
Casual Accidental Photo — Realistic AI Snapshot Prompt
Create hyper-realistic, flawed photos that look like spontaneous phone snapshots, with natural motion blur and uneven lighting for authentic, candid results.
1 / 2
提示词 · EN
ChatGPT, we’ve been together for quite a while now. I’d like to see what you look like. Please generate an image that feels like a casual snapshot taken on an argcamera: iPhone—nothing with a clear subject or intentional composition. Just something very ordinary, even slightly flawed, like an accidental photo. The image should have argvisual artifacts: a bit of motion blur, uneven lighting, and be slightly overexposed. The angle should feel awkward, the framing messy, and overall it should give off a “too real, spur-of-the-moment” vibe like a argscenario: selfie accidentally taken while pulling a phone out of a pocket.
关于这个提示词
Create hyper-realistic, flawed photos that look like spontaneous phone snapshots, with natural motion blur and uneven lighting for authentic, candid results. 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 camera、visual artifacts、scenario 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 this prompt into your preferred AI image generator (like Sora 2, DALL·E, or Midjourney).
- Select a model optimized for photorealism, such as GPT-4o, Sora 2, or SDXL.
- Customize the variables within the curly brackets (
{}) to change the camera, visual flaws, or scenario. - Adjust any specific settings in your generator (e.g., aspect ratio) to match a typical phone photo.
- Generate the image and regenerate if it looks too polished; the goal is imperfection.
