Alibaba Announces Upcoming Open-Weights Release of 2.4T Parameter Qwen3.8-Max Model
Alibaba's Qwen team has announced the upcoming open-weights release of Qwen3.8-Max, a massive 2.4-trillion-parameter model, alongside the smaller Qwen3.8-27B model. Built on the Qwen3.5 architecture, Qwen3.8-Max features advanced autonomous coding capabilities and adjustable reasoning depth. The release of a 2.4-trillion-parameter open-weights model marks a significant milestone for the open-source AI community, offering unprecedented local capabilities. It challenges proprietary frontier models by providing developers with access to top-tier coding and reasoning performance. Qwen3.8-Max supports a `reasoning_effort` parameter (with xhigh, medium, and low settings) to balance accuracy, speed, and cost. Additionally, the model is capable of autonomous coding, allowing it to complete complex, multi-day projects from scratch with minimal human intervention.
## BACKGROUND
Qwen is a family of large language models developed by Alibaba Cloud, known for strong performance in multilingual tasks and coding. Open-weights models allow developers to download and run the neural network's weights locally, providing greater control and privacy compared to API-only proprietary models.