MINIMAX BUILDING 2.7T PARAMETER MODEL FOR Q3 RELEASE

M3 Pro would surpass existing Chinese AI models in scale. The 2.7 trillion-parameter size positions it among the largest language models in development globally. MiniMax's decision to open-source the model represents a significant move in the competitive AI landscape, where major developers have increasingly restricted access to their largest models. Open-sourcing allows broader adoption and development by the research community. The Q3 timeline suggests the model is in advanced development stages. MiniMax, which has previously released smaller models and applications, continues expanding its AI capabilities to compete with larger players in both China and internationally. Details on M3 Pro's specific capabilities, training data, and computational requirements remain limited. The release would mark a notable addition to the growing pool of large open-source language models available to developers and researchers.