What about Kimi and GLM?
replyThese are well behind the general state of the art (1yr or so), though they're arguably the best openly-available models.
replyIdk man, GLM 5 in my tests matches opus 4.5 which is what, two months old?
replyAccording to artificial analysis ranking, GLM-5 is at #4 after Claude Opus 4.5, GPT-5.2-xhigh and Claude Opus 4.6 .
reply