📝 Summary

When you plug the Daidalos-type everyday prompts for the illustrious-series models into Grok, you get "the Daidalos type itself" (see the attached image). Even if you feed it into ChatGPT or Nano Banana, or even into Imagen4, the illustration/phot realism ratio is ...

🎨 Chinese Prompt

把用于日常使用的 illustrious 系模型的代达罗斯型提示词放进 Grok,会直接出来“代达罗斯型本身”(附图)
就算放进 ChatGPT、Nano Banana,甚至 Imagen4,除去插画/写实的比例差异,结果几乎相同。

代达罗斯型的提示词,可能是个非常稳定且强大的东西? https://t.co/O03LupSVKj

🎨 English Prompt

When I put a prompt for a Daedalus-type “illustrious” model meant for everyday use into Grok, it outputs “Daedalus itself” (attached image).

Even if I input it into ChatGPT or Nano Banana, or even Imagen 4, the result is almost the same aside from the illustration-to-realism ratio.

Maybe the Daedalus-type prompt is an extremely stable, strong one? https://t.co/O03LupSVKj

📷 Images

ダイダロス型の普段使いのillustrious系モデル用プロンプトをGrokに突っ込むと、「ダイダロス型そのもの」が出てく...