I've been running some of the biggest open-weight LLMs for free on Nvidia's cloud
…For example, MiniMax M2.7 is there, which I've actually tested in the past with a MiniMax subscription given that it's one of Claude's most credible open-weight competitors…