Chinese AI Developer MiniMax Builds 2.7T-Parameter M3 Pro Model for Q3 Release
A report from The Information says MiniMax is preparing a 2.7 trillion-parameter model called M3 Pro and plans to open-source it as early as Q3.
711K5714782.6K
“Huge: China’s MiniMax Plans to Launch 2.7-Trillion Parameter Model (MiniMax Pro) tl;dr MiniMax is preparing a 2.7T parameter open-source model, potentially launching as early as Q3.”
KimmonismusTECH#1364Story Brief
The Information reported that Chinese AI company MiniMax is preparing a 2.7 trillion-parameter model, internally called M3 Pro, and could release it as early as Q3. Posts on X quickly seized on the scale claim, with one widely shared summary from @kimmonismus arguing the model would be larger than any Chinese model currently on the market and more than six times the size of MiniMax's current M3.
“Minimax needs to flip the narrative. With 2.7T it should be possible. Go on, surprise me. Kimi did it, Zhipu did it. Can you?”
TeortaxesTexRelated Stories
4 linkedExclusive: China’s MiniMax Plans to Launch 2.7-Trillion Parameter ModelThe report behind the Q3 timing, model name, and 2.7T claim.The InformationMiniMax M3: Open-weight model with a million-token context challenges proprietary leadersExplains how MiniMax has been positioning its current M3 model on coding and long-context performance.The Decoder
