Alibaba Group is expanding its push into enterprise artificial intelligence, complementing its recently launched agent platform Wukong with a new flagship language model, CNBC reports. 

Debut of the new flagship model

On April 20, Alibaba Group released Qwen3.6-Max-Preview, a preview version of its most powerful language model to date. Compared to the recently launched Qwen3.6-Plus, the new model shows improvements in agent-based coding, world knowledge, and instruction following.

It expands the rapidly growing Qwen3.6 family, which also includes Qwen3.6-Flash and the open-source Qwen3.6-35B-A3B model.

Wall Street focuses on AI investment push

Analysts have highlighted the acceleration of Alibaba Group’s AI spending. According to Bernstein, the company’s AI investments nearly doubled in the March quarter to around 20 billion yuan, or approximately $2.93 billion.

Management has set a five-year target to exceed $100 billion in combined annual revenue from cloud computing and AI services. This is backed by a three-year commitment to invest at least 380 billion yuan (around $53 billion) in AI infrastructure and cloud computing.

Bank of America maintains Buy rating

Bank of America reaffirmed its Buy rating with a $180 price target. Analyst Joyce Ju noted that the scale of reinvestment had been “significantly underestimated.”

Goldman Sachs upgraded Alibaba Group to Conviction Buy, forecasting a recovery in earnings during fiscal years 2027–2028 driven by AI and cloud growth.

Alibaba Group shares, traded in Hong Kong, rose more than 14% in April through mid-month, putting them on track for their strongest monthly gain since January.