https://www.reddit.com/r/LocalLLaMA/comments/1n2ghx4/ama_with_zai_the_lab_behind_glm_models/?sort=qa 
Q: What do you think open weights models like GLM4.5 or Kimi K2 are doing differently to closer frontier commercial models like GPT-5, Gemini, Claude etc., and what needs to change in order to catch up or overtake those closed models? Will it ever happen? 
您认为 GLM4.5 或 Kimi K2 等开放式重量模型与 GPT-5、Gemi...