Qwen4-Max is here!
@Alibaba_Qwen has released its new flagship AI model with 2.4 trillion parameters, native vision and a 1 million token context window. Its mixture-of-experts architecture activates 71 billion parameters per token, 25% fewer than its predecessor.
Alibaba says the model is stronger at coding, agent tasks and visual reasoning. On its own Work360 benchmark, Qwen4-Max used 35% fewer tokens and reduced the average cost per task by around 60% compared to Qwen3.8-Max.
Qwen4-Max is available through QwenCloud and Alibaba's apps. Smaller open-weight Qwen4 models are also coming, though no release date has been announced.