MiniMax-Text-01
MiniMax · 2025-01-15 · 456.1B parameters
MiniMax Text-01 is MiniMax's 456-billion-parameter open-weight language model with approximately 45.9 billion active parameters, released in January 2025. It uses a hybrid architecture combining Lightning Attention, softmax attention, and Mixture-of-Experts layers, supports up to 1-million-token training context and 4-million-token inference context, and serves as the foundation language backbone for the MiniMax-01 model family.