At work, I get to use frontier models like Anthropic’s Claude Sonnet 4. I’ve been messing around with Z.ai’s GLM-4.5, Kimi K2 and xAI’s Grok 4 at home …
I was playing around with Chinese Large Language Models (LLMs). Both Kimi K2 and GLM-4.5 feel a bit slower than Claude Sonnet 4 when I’m using Claude …