MiniMax M3 is a multimodal foundation model with 512K token context, native multimodal understanding, and MiniMax Sparse Attention (MSA) for efficient long-context inference.
View detailed pricing and capabilities for this provider.
langrouter/minimax-m3