Qwen 3.8 Max: 2.44 Trillion Parameters, Open-Weight AI

Qwen 3.8 Max has launched as the largest open-weight multimodal AI model, featuring an impressive 2.44 trillion parameters. This new model includes Recursive Self-Improvement, which allows it to optimize its own processes over time. For example, it enhanced operational frameworks during a 16-day refinement cycle. Its open-weight design aims for transparency and accessibility for developers and researchers. The model shows versatility in areas like Python execution, structured reasoning, and creative tasks such as animation design. It also performs in real-world applications like space tracking and interactive mapping. An upcoming Qwen 3.8 27B model is designed to extend advanced AI capabilities to consumer devices. This development could make advanced AI more accessible to a wider range of users.