Aug 20, 2026 earnings call — AI and cloud commercialization drive growth
Total revenue rose 9% to RMB 269 billion, led by 45% cloud growth and AI-related products with RMB 49.5 billion ARR.
Revenue $39.6B (est $39.5B) · EPS $1.25 (est $1.94)
Key quotes
"This quarter, Alibaba Cloud's external revenue growth accelerated to 45%, a 22 quarter high, while adjusted EBITDA margin reached 11.6%."
"Last week, we opened the modeled weights of Qwen 3.8 Max with 2.4 trillion parameters and the Qwen 3.8-27B model series."
"The Quick Commerce business is expected to achieve overall profitability in FY '29."
Analyst Q&A 5
CapEx increase and payback
Q (Alicis a Yap · Citi): Why did CapEx rise so much this quarter? What will CapEx look like in coming quarters, any update to the RMB 380 billion 3-year plan, how is CapEx split by service, and what return do you expect?
A: They said the 3-year plan is RMB 380 billion and as of end of June they had spent RMB 190 billion. This quarter’s higher spending (RMB 67.1 billion) mainly reflects volatility in hardware delivery schedules, plus increased CPU procurement because of strong demand and higher semiconductor prices. They said AI CapEx can breakeven in about 3 years and could shorten to 2.5 or 2 years as margins and proprietary chip use improve.
"So the increase primarily reflects volatility in those equipment delivery schedules." — Unknown Executive
E-commerce group strategy and quick commerce
Q (Charlene Liu · HSBC): After regrouping e-commerce lines, what are the strategic focuses, and what is the update on quick commerce?
A: They said the e-commerce business is now organized into 4 core areas: China e-commerce, quick commerce, international e-commerce and B2B global wholesale. For China e-commerce they will strengthen supply capabilities, support brands and industrial-cluster suppliers, and use AI to improve shopping and merchant operations. For quick commerce they will integrate Freshippo and Tmall supermarket, expand front warehouses, expect nonfood to pass food next fiscal year, and target overall quick commerce profitability in FY '29 and a long-term contribution of 30% of platform GMV; international and B2B areas will also focus on profitability and AI agents (Accio Work had over 50,000 paying merchants).
"Moving forward, we'll be updating progress on 4 core areas: China e-commerce, quick commerce, international e-commerce and global -- B2B global wholesale." — Unknown Executive
Cloud growth pace and drivers
Q (Yang Bai · CICC): How fast will cloud growth run in coming quarters and what are the core short- and long-term drivers?
A: They said external revenue for AI and Cloud has accelerated for 9 consecutive quarters and reached 45% growth this quarter, and they expect growth to keep accelerating in coming quarters. They reported AI-related products generated RMB 12.4 billion this quarter (annualized USD 7.3 billion) and forecast next quarter’s annualized AI revenue will approach USD 10 billion, with EBITDA margin improving sequentially. Long term, they cited full-stack investments (chips, cloud infrastructure, models), scale and network effects, and said they are confident in reaching RMB 100 billion external cloud revenue by 2030 and achieving about 20% gross margin.
"We expect revenue growth to continue accelerating over the coming quarters." — Unknown Executive
MaaS ARR and model mix
Q (Yuan Liao · CITICS): Will you change the year-end RMB 30 billion MaaS ARR target? What share is proprietary models vs third-party, and how will open models affect margins?
A: They said ARR had surpassed RMB 16 billion as of August and they remain confident of reaching the year-end target of RMB 30 billion ARR. They said proprietary models still account for the majority of MaaS revenue, but third-party model revenue is also meaningful and customers often use multiple models. They added that gross margin from hosting proprietary models versus third-party models on the MaaS platform is highly comparable, and a thriving open-model ecosystem is favorable for the cloud provider.
"So given the current growth momentum as well as the pipeline of new models slated for launch, we remain confident that we will achieve our year-end target of RMB 30 billion ARR by the end of the year." — Unknown Executive
Where value will sit in full-stack AI
Q (Alex Yao · JPMorgan): Which layer (chips, cloud infra, models, orchestration) will capture most value over time? If compute/hardware holds value, how should we think about competition and government allocation?
A: They said they are investing across the full stack so they can capture value no matter which layer concentrates it, and that there is inherent uncertainty about which layer will dominate long term. In the short term they believe most value will be in chips and AI cloud infrastructure, and they described T-Head chips as already deployed at scale (previous generation over 500,000 shipped) serving hundreds of external customers, with generation-2 development starting in the second half of this year. They also said API monetization is a transitional model and the ultimate monetization will be delivering products and results as models approach AGI.
"I think that most of the value will be in chips and in AI cloud infrastructure." — Unknown Executive