Qwen releases the open Qwen3.8-2.4T-A95B
QwenCloud lists Qwen3.8-2.4T-A95B as the open version of the flagship Qwen3.8 family.
What happened
Event details
Alibaba's QwenCloud lists Qwen3.8-2.4T-A95B as an open flagship Qwen3.8 model. It uses sparse MoE and hybrid attention, with 2.4 trillion total parameters and about 95 billion active per step, supports one million tokens of context, enables thinking by default, and targets coding, productivity, research and long-horizon agent tasks.
Assessment
Why it matters
Qwen3.8-2.4T-A95B formally adds an open model to the flagship Qwen3.8 release record, using an extremely large sparse-MoE design for coding, research and long-horizon agent workloads.
Availability
Access notes
QwenCloud's official release record lists it as an open version, but does not include the license text or a weight-download link.