Alibaba Triples Profit to $9.6 Billion as AI Cloud Revenue Jumps 53%
Alibaba tripled its net profit to 64.4 billion yuan ($9.6 billion) for the quarter ending in September, but investment income helped as much as the core business did. The clearest sign of real AI demand is the cloud unit, where companies rent servers and chips for neural networks. Segment revenue rose 53% to 61 billion yuan. Meanwhile, money spent on data centers is leaving faster than it comes in. The takeaway: renting AI capacity is already profitable, while chatbots like Qwen remain expensive to run.
AI-processed from TNW; edited by Hamidun News
Imagine a startup whose quarterly profit tripled while its cash balance is draining faster than ever. That is what Alibaba's quarterly report looks like, and it shows well where the real money in the AI boom is and where there are still only costs. The company reported results for the quarter ending in September. If you read only the headlines, everything looks triumphant. If you read the numbers, the picture is more complicated.
Where the tripled profit came from
Alibaba's net profit tripled to 64.4 billion yuan (about $9.6 billion). Analysts polled by FactSet had expected roughly 23.7 billion yuan, so the result beat the forecast by almost threefold. Revenue rose only 10% to 271.5 billion yuan ($40.5 billion), beating expectations by less than 1%. The company's US shares even dipped slightly in pre-market trading.
The gap is explained by higher net investment income and operating income driving profit. It is as if your earnings grew not only from sales but also from well-timed bets on other companies. For assessing the health of the business, the lines tied to AI matter more.
AI cloud is already earning money
The AI Cloud and Compute Services segment (which includes the chip unit T-Head) essentially rents out servers, chips and ready-made AI services. Its revenue grew 53% to 61 billion yuan. Recurring revenue from external customers, meaning companies outside Alibaba's ecosystem, rose 67%. AI-related products have been growing at triple-digit rates for the 13th consecutive quarter and brought in 17.9 billion yuan.
Importantly, the segment is not only growing but also earning money: adjusted EBITA (earnings before interest, taxes and amortization) rose 121% to 8.8 billion yuan. For a developer, this is a signal: the business of selling compute for neural networks works not only for American clouds.
Where the money leaks out
The neighboring AI Labs and Applications unit, which develops the Qwen models, increased revenue by 51%. But its adjusted loss tripled to 10.5 billion yuan. Alibaba attributes this to rising AI spending and the cost of running the Qwen app. It is a familiar story: every user query to a chatbot costs money in compute, and until revenue covers those expenses, the losses persist.
Capital expenditure rose 60% to 50.5 billion yuan as the company builds AI infrastructure. Free cash flow, meaning the money left after all expenses and investments, was negative: an outflow of 34.5 billion yuan versus 21.8 billion yuan a year earlier. Chief Financial Officer Toby Xu promised "disciplined investment" focused on return on invested capital and a goal of reaching positive free cash flow. He did not give a timeline.
Meanwhile, the legacy business is slowing. Revenue from merchant services, mainly commissions on Taobao and Tmall, fell 5%. Adjusted profit for the e-commerce group still rose 243%, because losses at the Taobao Instant Commerce quick-delivery service shrank.
What this means in practice
First, if you are choosing a cloud or building a product on open Qwen models, the provider is clearly not pulling back on investment: money going into infrastructure is growing at 60% a year.
Second, the Qwen unit's loss is a reminder that a mass-market free AI app turns into a line item in the cost column. Those building their own AI products should calculate the cost of serving one user in advance and should not assume that traffic will turn into money on its own.
Third, geopolitics is part of the backdrop. The report came out a few days after the US move to restrict Chinese companies' access to advanced chips. According to Bloomberg, Alibaba's market value is now about $200 billion, less than half its level from a little over a year ago. Alibaba also confirmed that one of its products will reach commercial release in the first quarter of 2027, and this week it updated its flagship model.
Bottom line: Alibaba's renting out of AI capacity is already producing profit, while the race for models and users is still being paid for out of pocket. Whether the investments pay off before investor patience runs out will become clear in the coming quarters.
- Nov 21, 2026Eight Open-Source Tools for AI Agents: Browser Use Leads with 121,000 Stars
- Nov 21, 2026Thinking Machines Releases Open-Weight Inkling 2 Mini with 26.4B Parameters for 32 GB Macs
- Nov 21, 2026MLCommons Starts Measuring AI APIs Under Real Load: 7 Runs for Agents
- Nov 21, 2026Unsloth Updates Compressed Qwen3.8-27B: Accuracy Up 5–10% at the Same Size
Key Questions Answered
Where the tripled profit came from?
Alibaba's net profit tripled to 64.4 billion yuan (about $9.6 billion). Analysts polled by FactSet had expected roughly 23.7 billion yuan, so the result beat the forecast by almost threefold. Revenue rose only 10% to 271.5 billion yuan ($40.5 billion), beating expectations by less than 1%. The company's US shares even dipped slightly in pre-market trading. The gap is explained by higher net investment income and operating income driving profit.
What are the key details on AI cloud is already earning money?
The AI Cloud and Compute Services segment (which includes the chip unit T-Head) essentially rents out servers, chips and ready-made AI services. Its revenue grew 53% to 61 billion yuan. Recurring revenue from external customers, meaning companies outside Alibaba's ecosystem, rose 67%. AI-related products have been growing at triple-digit rates for the 13th consecutive quarter and brought in 17.9 billion yuan.
Where the money leaks out?
The neighboring AI Labs and Applications unit, which develops the Qwen models, increased revenue by 51%. But its adjusted loss tripled to 10.5 billion yuan. Alibaba attributes this to rising AI spending and the cost of running the Qwen app. It is a familiar story: every user query to a chatbot costs money in compute, and until revenue covers those expenses, the losses persist. Capital expenditure rose 60% to 50.5 billion yuan as the company builds AI infrastructure.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.