Alibaba Qwen3.8 Foundation Model

Overview

Around August 15–19, 2026 Alibaba released Qwen3.8, a new-generation Tongyi Qianwen foundation model whose flagship Qwen3.8-Max reaches about 2.4 trillion total parameters, 95 billion activation parameters and a 1,000,000-token context window with native visual input support.[6][14] The model is designed for fast response, high-quality outputs and easier deployment; after quantization it runs on consumer-grade GPUs, making it attractive for on-premise and edge deployments.[6] This positions it as a competitive open-weight option in the global frontier model ecosystem, especially for Chinese-language and enterprise workloads.[6][14]