MiniMax-M1 open-weight, large-scale hybrid-attention reasoning model from Hacker News on 2025-06-18 06:53 (#6Y2A9) Comments
