What's the best model for this prompt?

Sora 2 or similar image-generation models work best for achieving the desired low-fidelity, pixelated aesthetic.

Can I adjust the level of 'roughness' in the doodle?

Yes, you can modify the prompt to emphasize or reduce elements like resolution, line quality, or color palette.

Will the output be recognizable from the original image?

Yes, the prompt ensures the image remains vaguely identifiable, but with intentionally awkward, confusing details.

MS Paint Style Doodle — Sora 2 Prompt

Transform any image into a clumsy, low-res MS Paint-style doodle that humorously captures the essence of 'soul mouse drawing.'

提示词 · EN

用尽可能粗糙、潦草且糟糕的风格重绘这张图片。背景应为白色，营造出一种“灵魂鼠标画”的感觉，就像是用电脑鼠标在系统自带的绘图软件中画出来的一样。它应该能让人模糊地辨认出原图，但又显得有些不对劲——笨拙、令人困惑，并带有低分辨率的像素质感，以强调那种惨不忍睹的效果。好吧，随便了，你想怎么画就怎么画。

关于这个提示词

Transform any image into a clumsy, low-res MS Paint-style doodle that humorously captures the essence of 'soul mouse drawing.'。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。