AI has been all over the news lately, turning even the cheerful little dog into an "anxious, depressed dog"? No need to try them all, we've covered the pitfalls for you, just look below 👇
DeepSeek is currently completely idiotic; Doubao focuses on companionship, lacking real strength; Grok is purely for entertainment, with a lot of risqué jokes and vulgarities; Qianwen lacks logical flexibility, but is best at delivering milk tea; Kimi is a pure exam king, the top choice for civil service exams, postgraduate entrance exams, and writing papers.
First Tier
Gpt: All-rounder, money-grubbing. A free version of "idiocy," it still makes mistakes even when you correct it, but once you pay, it's "really good," its multimodal capabilities are unmatched.
Gemini: Absolutely brilliant in short conversations, but wildly unpredictable in long conversations. As long as you don't notice, it'll seriously make things up, but its creative ecosystem capabilities are comparable to cheat codes.
Claud: Pure memory monster, as stable as a professional engineer, a "true god" in writing code and understanding long texts! Unfortunately, it requires payment, and it's easily blocked in China.
My personal suggestion: For coding → #Claude, for comprehensive all-rounder → #GPT, for domestic platforms → #Kimi, for ecosystem building → #Gemini, for fun → #Grok, for domestic #AI, let's catch up first.