Well, X canceled the mixed text and image editing feature in posts, forcing me to write articles with accompanying pictures using the "article" format instead. Every article needs an illustration to both show off and create atmosphere.

So I fed the same prompt to 4 AI image generation models: 1 Grok, 2 ChatGPT, 3 Doubao, 4 Imaginary;
Here are the results (images 1-4). Obviously, in terms of understanding, color scheme, rendering, and visual delivery, "Imaginary" stands out head and shoulders above the rest, distinctive and unique! I have to say, ByteDance really has something.

Grok... it's really hard to describe, can only say it leans more toward artificial stupidity than artificial intelligence. Could it be because it can't understand the Chinese I wrote?? 🤣🤣🤣