你喜欢超长的提示词吗？❓

Do you like BIG PROMPTS?❓ I mean REALLY BIG AI VIDEO prompts! Wonder what a 12,000 character one would be like? Wonder no more! Pretty neat! Would love to see others results! Works well with Seedance 2.0 Will likely overwhelm any other model. This is VERY experimental. Behold the results of an over 13,000+ one below: I spent some more time on the BIG PROMPT HACK I had posted about before but at its core: Paste your text prompt into GPT5.5 "Create a high resolution 16x9 PNG of this text prompt, clear black text against white background. Ensure all text is retained and readable and cleanly organized" You can also do this in other programs manually such as CANVA. Depending on how high resolution the image is you may have to compress it a bit as different platforms will allow various maximum image uploads. Upload the prompt image as your omni ref. Upload any supporting image or omnirefs as needed. You can (and should) add supportive prompting to highlight key prompt needs. Supportive prompting example used for this: Use Image 1 as the prompt and visual reference for the scene composition, subjects, and environment. Use Image 2 as the style reference for color grading, lighting quality, and overall aesthetic treatment. High-energy real footage aesthetic with dynamic handheld camera work, natural motion blur, and authentic documentary-style framing. Raw, unpolished cinematography with visible camera shake and reactive movements. Vibrant, saturated color palette with punchy contrast. Fast-paced action with rapid subject movement and aggressive camera tracking. Gritty, street-level realism with natural

你喜欢超长的提示词吗？❓ 我指的是真正超长的 AI 视频提示词！ 想知道一个 12,000 字符的会是什么样的？ 不用再猜了！ 挺酷的！很想看看其他人的成果！ 与 Seedance 2.0 配合使用效果很好。 很可能会让其他任何模型不堪重负。 这是非常实验性的。 请看下方这个超过 13,000+ 字符的结果： 我花了一些时间继续研究我之前发过的那个超长提示词技巧，但核心内容是： 把你的文字提示词粘贴到 GPT5.5 中，输入："Create a high resolution 16x9 PNG of this text prompt, clear black text against white background. Ensure all text is retained and readable and cleanly organized" 你也可以在其他程序中手动完成，比如 CANVA。 根据图片分辨率的高低，你可能需要稍微压缩一下，因为不同平台允许上传的最大图片大小各不相同。 把这个提示词图片作为你的 omni ref 上传。根据需要上传任何支持图片或 omnirefs。 你可以（而且应该）添加辅助提示来突出关键的提示需求。本次使用的辅助提示示例如下： Use Image 1 as the prompt and visual reference for the scene composition, subjects, and environment. Use Image 2 as the style reference for color grading, lighting quality, and overall aesthetic treatment. High-energy real footage aesthetic with dynamic handheld camera work, natural motion blur, and authentic documentary-style framing. Raw, unpolished cinematography with visible camera shake and reactive movements. Vibrant, saturated color palette with punchy contrast. Fast-paced action with rapid subject movement and aggressive camera tracking. Gritty, street-level realism with natural