Chinese AI company MiniMax is preparing to release a new large language model named M3 Pro, which contains 2.7 trillion parameters. The model is expected to be released and open-sourced as early as the third quarter of this year.

  • M3 Pro features 2.7 trillion parameters, significantly larger than MiniMax's current flagship M3 model, which has 428 billion parameters.
  • The new model aims to provide significant improvements in handling complex reasoning and multi-step tasks.
  • Larger-scale models are noted for being more capable of processing complex reasoning and multi-step instruction-based tasks.

The release of this larger model is intended to enhance MiniMax's capabilities in processing complex instructions.