Serve Qwen3.8-2.4T-A95B, a 2.4T-Parameter Model, with Configurable Reasoning on NVIDIA GB300 NVL72
Alibaba releases open weights for 2.4T-parameter Qwen3.8-Max, its largest open-weight model
Alibaba has released open weights for Qwen3.8-2.4T-A95B (Qwen3.8-Max), a 2.4T-parameter mixture-of-experts model with 95B parameters activated per token, bringing near-frontier capabilities to the open ecosystem. The model features hybrid full and linear attention and runs on NVIDIA GB300 NVL72 hardware. This is a significant signal for the open-weight frontier race, as it closes the gap between proprietary and open models at massive scale.