Journalism begins where hype ends

,,

The rise of powerful AI will be either the best, or the worst thing, ever to happen to humanity. We do not yet know which."

—Stephen Hawking

Alibaba’s Qwen Team Introduces Qwen4-Max, Targets Complex AI Workflows

Alibaba's Qwen team has launched Qwen4-Max, its latest flagship artificial intelligence model, featuring a 2.4-trillion-parameter backbone, native vision capabilities and a one-million-token context window. The company claims the model reduces average task costs by 62% compared with its predecessor in its internal Work360 evaluation.
Qwen4-Max AI model launch announcement by Alibaba's Qwen team
November 16, 2026 12:04 PM IST | Written by Vaibhav Jha

Alibaba’s Qwen team released on Monday its latest flagship artificial intelligence model Qwen4-Max of the Qwen4 series, with a 2.4-trillion-parameter backbone, native vision capabilities, and a one-million-token context window.

Qwen4-Max has been positioned by the Qwen team as an advanced “AI teammate” that is optimized for real-world productivity, complex agentic workflows, coding, frontend 3D design as well as long-horizon task execution.

According to the Qwen team, the model uses a new sparse Mixture-of-Experts (MoE) architecture with 2.4T backbone parameters, 71B activated per token, and an additional 328B N-gram Embedding parameters.

 

“The architecture systematically upgrades the sparse MoE model across attention, residual connections, and embeddings, improving computational efficiency, model capacity, and training stability,” read a statement from Qwen.

Team Qwen has also claimed that Qwen4-Max outperforms Qwen3.8-Max with fewer tokens on their company’s internal benchmark framework “Work360” where token usage has dropped by approximately 24% and 35% at Xhigh and High effort, respectively. The company informed that the Work360 framework has been developed for strengthening the model’s collaborative capabilities across industrial and professional workflows.

“In the latest evaluation, Alibaba Qwen4-Max scored 52.69, up from 47.74 for Qwen3.8-Max, marking a substantial improvement over the previous generation. Qwen4-Max combines broader industry coverage with stronger professional task execution and collaboration, making it a more capable and versatile tool for real-world productivity. We have also lowered API prices compared with Qwen3.8-Max, with particularly large reductions in cached-input pricing. In our Work360 evaluation, the average cost per task drops from $4.43 to $1.67 — a 62% reduction,” read a statement from Qwen.

Also Read: Alibaba Launches ‘Most Capable’ Qwen3.8-Max, Set to Become First Open-Weight Max Model Next Week

Author

  • Vaibhav Jha, editor and co-founder at AI FrontPage

    Vaibhav Jha is an Editor and Co-founder of AI FrontPage. In his decade long career in journalism, Vaibhav has reported for publications including The Indian Express, Hindustan Times, and The New York Times, covering the intersection of technology, policy, and society. Outside work, he’s usually trying to persuade people to watch Anurag Kashyap films.

    LinkedIn