Alibaba’s open-weight Qwen3.8-Max takes on long-horizon AI tasks with 2.4 trillion parameters
Alibaba is releasing Qwen3.8-Max, an open-weight model with 2.4 trillion parameters designed for long-horizon autonomous tasks like chip design and research reproduction.
Alibaba Cloud is preparing to release the weights for Qwen3.8-Max, a flagship large language model featuring 2.4 trillion parameters. The model is specifically engineered for long-horizon reasoning and autonomous task execution over extended periods.
According to the development team, the system can operate independently for days to complete complex workflows. Demonstrated use cases include reproducing academic research papers and autonomously designing integrated circuits, highlighting its potential as an advanced AI agent.
The move to open-weight distribution aligns with recent trends in the LLM sector, where major providers are sharing larger models to foster ecosystem development. This release positions Alibaba competitively against other frontier model providers focusing on agentic capabilities.
The weights are scheduled for public release next week. Industry observers will likely assess the model's performance on benchmark tasks to determine how its scale translates into practical utility for enterprise and research applications.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.