Shares of Alibaba surged 4.8% to $128.06 after the company launched its largest-ever artificial intelligence model — a system capable of understanding text, images, and video simultaneously — extending a rally that has added roughly 11% over the past week. Hong Kong-listed shares rose about 6% , as investors bet the new model strengthens Alibaba's position atop China's fast-growing cloud market. But the real question isn't whether the technology is impressive — it's whether Alibaba can turn AI excitement into earnings before its cash pile erodes further.
The Model That Sparked the Rally Competes With the World's Best
Alibaba described the new system as the most capable in its lineup, with 2.4 trillion total parameters and 95 billion active parameters.
Benchmark comparisons showed it performing at comparable levels to Anthropic's top model on several coding and general tasks, and scoring above it on some document and multimodal benchmarks. That matters because perception of technical leadership directly influences which cloud provider enterprises choose. However, no independent third-party evaluator has scored it yet, and the "open-weight" promise — letting outside developers download and run it — remains unfulfilled so far.
A Billion Downloads Give Alibaba a Built-In Customer Pipeline
Alibaba's AI model family has officially crossed one billion cumulative downloads, overtaking Meta's Llama to become the world's most widely used open-source AI system.
The company's strategy focuses on offering powerful, low-cost models to drive mass adoption, then converting that developer base into paying cloud customers.
The customer base for its model service platform grew eightfold year-over-year. That pipeline is the bridge between free downloads and real revenue.
Cloud Is Booming, but Profits Are Collapsing
Cloud revenue jumped 38% year-over-year to about $6 billion last quarter, with AI-related products making up 30% of all cloud revenue and growing at triple-digit rates. Yet adjusted operating profit plummeted 84%, with the company-wide margin compressing to just 2%.
Free cash flow swung to a $6.8 billion outflow for the full fiscal year , fueled by heavy infrastructure and commerce spending. Investors cheering AI progress must weigh whether growth will eventually generate profits — or simply consume them.
A $100 Billion Target Hangs Over Everything
CEO Eddie Wu has outlined an ambitious $100 billion five-year external cloud and AI revenue target , and this model launch is designed to accelerate that trajectory. Annualized AI-related product revenue has already surpassed roughly $5.2 billion. The gap between $5 billion and $100 billion is enormous, but today's rally shows the market is willing to pay for the narrative — for now.