Alibaba has released Qwen 3.8-Max, a new large language model with 2.4 trillion parameters designed to excel at software engineering and long-duration tasks. The model uses a Mixture of Experts architecture, which routes each request to specialized components rather than activating the entire model at once. This approach improves efficiency and performance for specific workloads.
The model will become available next week through Alibaba Cloud's Model Studio platform. Qwen 3.8-Max is now the second largest open-source model by parameter count, behind only Moonshot's Kimi K3, which has 2.8 trillion parameters. The release marks Alibaba's continued push into competitive large-scale AI development alongside other major players in the space.
