Alibaba Group has introduced a new wave of AI technologies, including its latest large language model and a high-performance AI chip, aiming to establish itself as the central ‘AI factory’ in China amidst rising adoption of autonomous agents.

  • Qwen3.7-Max designed for complex reasoning and long-horizon autonomous tasks
  • Zhenwu M890 chip delivers triple the performance of earlier AI processors
  • Alibaba operates across all five layers of the AI stack in China

What happened

Alibaba Group revealed a suite of new AI products at the Alibaba Cloud Summit held in Hangzhou, highlighting its ambition to become China’s premier 'AI factory'. The company launched Qwen3.7-Max, a proprietary large language model optimized for agentic coding, intricate reasoning, and extended-duration tasks requiring sustained autonomous decision-making. Alibaba also introduced the new Zhenwu M890 AI training and inference chip, which offers three times the processing power of its predecessor, designed to support demanding AI workloads.

Senior Vice-President Liu Weiguang described Alibaba’s vision as pioneering a new manufacturing paradigm based on AI technology, centered on what the company calls 'training and inference factories.' Alibaba emphasized its unique position in China as an integrated AI and cloud provider operating across all five key layers of the AI stack — including chips, cloud infrastructure, models, service platforms, and agentic applications.

Why it matters

Alibaba’s rollout of an advanced large language model alongside custom-designed AI chips marks a significant step for China’s domestic AI development and commercialization. Qwen3.7-Max’s ability to autonomously handle complex, multi-step tasks for up to 35 hours without degrading performance signals Alibaba’s focus on durability and reliability in real-world AI applications, potentially outpacing many local competitors in domestic benchmarks.

The new Zhenwu M890 chip’s enhanced capabilities bolster Alibaba’s hardware foundation, crucial for supporting large-scale AI workloads and concurrent requests in agentic computing environments. By controlling the entire AI technology stack, Alibaba positions itself to compete more effectively against both domestic rivals and international players, although its models still trail leading US offerings in certain advanced capabilities.

What to watch next

Industry observers should monitor how Alibaba’s AI innovations influence the competitive landscape in China, particularly as the firm continues to ship tens of thousands of Zhenwu chips across various sectors, from automotive to finance. Scheduled future releases of next-generation chips like the Zhenwu V900 in 2027 and J900 in 2028 will be critical indicators of Alibaba’s ability to sustain momentum in AI hardware development.

Additionally, how Alibaba balances openness with proprietary control over its AI models will be telling. While smaller versions have been open-sourced, the flagship models remain closed amid rising commercialization pressures in China’s AI field. The company’s strategy for fostering developer adoption and integrating AI solutions across industries will shape its role as a leading AI factory going forward.

Source assisted: This briefing began from a discovered source item from SCMP China Tech. Open the original source.
How SignalDesk reports: feeds and outside sources are used for discovery. Public briefings are edited to add context, buyer relevance and attribution before they are published. Read the standards

Related briefings