Alibaba Announces Qwen 3 with 2.4 Trillion Parameters and Upcoming Open Weights
Alibaba's Qwen team has announced Qwen 3, a massive large language model featuring 2.4 trillion parameters. The team has also indicated that the open weights for this model will be released in the near future. A 2.4-trillion-parameter model represents a massive scale for open-weights AI, potentially rivaling the largest proprietary models and significantly advancing the open-source AI ecosystem. While the model has been announced, the exact release date and the hardware requirements for running a 2.4T parameter model locally have not yet been detailed.
## BACKGROUND
Qwen is a family of large language models developed by Alibaba Cloud, known for providing open-weight counterparts to proprietary models. In LLMs, parameters are the internal weights learned during training that determine the model's capacity to understand and generate text, with "open weights" meaning these parameters are publicly downloadable.