MiniMax Quater 3 or Open Sources 2.7T Parameter Model, Expected to Become the World's Largest Open Source Large Model

According to Beating Monitor, The Information reported citing sources familiar with the matter that MiniMax is developing a 27 trillion parameter large language model, which could be released as early as the third quarter and is planned to be open-sourced. The model is internally codenamed M3 Pro, but the final release name has not been determined.

If the weights are released as planned, this model is poised to become the world's largest open-source large model in terms of parameters. MiniMax's current flagship model, M3, has 428 billion parameters, and the new model is approximately 6.3 times the size of M3, surpassing several existing trillion-parameter open-weight models.

A larger parameter size is usually advantageous for complex reasoning and multi-step instruction tasks, but the final performance will still depend on training quality, activation functions, inference costs, and evaluation results.