Alibaba has unveiled Qwen4-Max, its newest flagship large language model in the Qwen4 family, and makes it generally available via its QwenCloud API. Multiple outlets say the company positions the model as a more capable “agent” or work-focused AI teammate, while also highlighting efficiency improvements and broad multimodal abilities.

Several reports center on pricing and efficiency. Outlets that cite Alibaba’s figures say Qwen4-Max costs $1.50 per million input tokens and $4.50 per million output tokens, and that average task costs drop by 62% versus Qwen3.8-Max in Alibaba’s Work360 evaluation. Other coverage focuses on the model’s technical design, including a mixture-of-experts setup where only part of the larger parameter set is active per token.

There are also differences in emphasis across publications. Some highlight that Qwen4-Max is closed and not available with open weights, while others stress how Alibaba frames it as a step forward for professional tasks. One outlet mentions training and attention optimizations as part of the performance and cost claims, while several focus on the reported million-token context window and native vision support.