kimi-k2 is pretty decent at coding but it’s nowhere near the SOTA models of Anthropic/OpenAI/Google.
Are you referring to the new reasoning version of Kimi K2?
Are you referring to the new reasoning version of Kimi K2?