On May 13, 2026, two of the most influential internet companies released their financial reports on the same day.

Alibaba and Tencent provided starkly different answers on the same AI exam. In the fourth quarter of FY2026, Alibaba Cloud's external commercialization revenue grew 40% YoY, with AI-related product revenue exceeding 30% for the first time. CEO Wu Yongming announced that Alibaba's full-stack AI technology investment had officially moved beyond the initial cultivation phase into a cycle of positive scale commercialization returns. Tencent reported Q1 2026 revenue of RMB 196.458 billion (+9% YoY) and operating profit of RMB 75.627 billion (+9% YoY). However, excluding the impact of new AI products like Hy, Yuanbao, and WorkBuddy, operating profit would have grown 17% YoY instead of 9%, meaning new AI products caused approximately RMB 8.8 billion in operating losses in a single quarter.

Pony Ma compared Tencent's AI journey to a maritime voyage at a shareholder meeting: 'A year ago, we thought we had boarded the ship, only to find it was leaking. Now we feel like we've stepped aboard but still can't sit comfortably. Tencent's foundational AI capabilities were not outstanding in the early stages, but through talent development, team management, and internal training in recent years, we are gradually entering a growth trajectory.'

On the same AI exam, Alibaba chose to become the 'super power plant' of the AI era, building infrastructure-level capabilities through full-stack self-developed technologies spanning chips, models, and MaaS platforms, with over 60% of its computing power serving external clients. Tencent, meanwhile, adhered to 'scenario-based dominance,' using AI to refactor (restructure) gaming, advertising, and office businesses through internal empowerment and application-first strategies.

These choices stem from their differing judgments of the AI industry and are inevitable decisions driven by their respective commercial DNA. Notably, beneath Alibaba's 'heavy investment' exterior, commercialization prospects are rapidly materializing, with AI model and application service ARR set to surpass RMB 10 billion, targeting RMB 30 billion by year-end. Tencent's restraint is not conservatism but a bet on long-term compounding gains from user scenarios.

There is no standard answer to which path is superior; together, they form two pillars supporting the upward growth of China's AI ecosystem.

Transparent Accounts vs. Hidden Strategies

The strategic divergence between Alibaba and Tencent is most vividly reflected in their financial disclosure approaches—one transparent, the other hidden. The presentation of numbers itself reveals their strategic intentions.

Alibaba's AI revenue is an open book, detailed to two decimal places. Alibaba Cloud Intelligence Group reported quarterly revenue of RMB 41.626 billion (+38% YoY), with a more critical structural breakthrough.

This quarter, Alibaba Cloud's external commercialization revenue accelerated to 40% growth, the fastest pace in nine quarters. AI-related product revenue exceeded 30% of cloud external commercialization revenue for the first time, reaching RMB 8.971 billion in quarterly revenue, achieving triple-digit YoY growth for the 11th consecutive quarter and surpassing RMB 35.8 billion in annualized recurring revenue. This marks Alibaba's first explicit disclosure of AI revenue proportion and annualized figures in its earnings report, clearly indicating that AI has upgraded from a 'growth variable' to a 'growth pillar.'

Wu Yongming further projected during the earnings call that AI model and application service ARR, including the BaiLian MaaS platform, is expected to surpass RMB 10 billion in the June quarter and reach RMB 30 billion by year-end, with AI-related product revenue proportion projected to exceed 50% in the coming year. This signifies that Alibaba Cloud's growth drivers will fully shift from traditional computing and storage to models, computing power, and Agent services, completing a structural revenue transformation.

Tencent's AI revenue remains a hidden account. In Q1, Tencent did not separately list AI revenue, but its pulling effect permeated growth drivers across multiple businesses.

Marketing services revenue grew 20% YoY to RMB 38.171 billion, with the earnings report citing 'upgraded AI-driven ad recommendation models' as a key growth driver. FinTech and enterprise services revenue grew 9% YoY to RMB 59.885 billion, with enterprise services revenue growing 20% YoY due to rising demand for AI-related services in domestic and overseas markets, driving cloud service revenue growth. Domestic gaming revenue reached RMB 45.4 billion, while international gaming revenue hit RMB 18.8 billion. Although AI's enhancement of user experience and engagement was not quantified, it was evident throughout.

However, the most intriguing figure in Tencent's earnings report lies not in revenue but in profit. Excluding the revenue, cost, and expense impacts of new AI products like Hy, Yuanbao, CodeBuddy, WorkBuddy, and QClaw, Tencent's Q1 Non-IFRS operating profit would have grown 17% YoY instead of the actual 9%. This 8-percentage-point growth gap corresponds to approximately RMB 8.8 billion in quarterly investment, meaning new AI products consumed RMB 8.8 billion in profits in a single quarter.

For comparison, Tencent's Q1 Non-IFRS operating profit totaled RMB 75.159 billion, with AI investment accounting for about 11.7%. Meanwhile, free cash flow reached RMB 56.7 billion, indicating that core businesses generate sufficient cash to cover AI investments with surplus. This massive investment has yet to translate into quantifiable AI revenue, instead accumulating in model capabilities, user habits, and scenario coverage—a 'hidden account' recorded beyond the balance sheet.

One transparent account, one hidden account. Alibaba separately lists AI revenue, quantifying it clearly to signal commercialization to the market—this is 'unsheathing the sword.' Tencent integrates AI into various businesses, letting the market perceive AI's penetration through implicit indicators like ad click-through rates and user engagement—this is 'building momentum.'

These two disclosure logics are not superior or inferior but reflect the companies' differing definitions of AI value. At this stage, Alibaba treats AI as a directly tradable, independently priced business; Tencent views AI as a capability deeply integrated into product textures, difficult to peel off (separate).

Genes Determine Paths, Timing Determines Pace

If earnings data reveal 'what is happening,' strategic DNA and commercialization timing explain 'why.' Specifically, why Alibaba and Tencent chose vastly different investment rhythms in AI.

Alibaba was born in e-commerce and matured through cloud computing. Taobao and Tmall's e-commerce foundation necessitated robust computing infrastructure, with annual Double 11 traffic surges serving as Alibaba Cloud's ultimate stress test. Meanwhile, its e-commerce DNA endowed Alibaba with another capability: B-end service expertise.

From 1688 to Alibaba Cloud, Alibaba's core clients have always been enterprises, giving it profound insights into B-end clients' willingness to pay and pain points. Wu Yongming explicitly stated during the earnings call that currently, both globally and in China, enterprises demonstrate stronger willingness to pay, with Alibaba allocating most of its inference resources to B-end commercialization.

Thus, when the AI wave arrived, Alibaba naturally extended its existing 'cloud + AI' trajectory—a smooth continuation rather than a radical pivot. Alibaba's investment stance is 'full throttle ahead,' with Wu Yongming stating that compared to 2022 (before the AI large model boom), future computing center scale must grow 'at least tenfold,' and AI infrastructure investment will far exceed the previously announced RMB 380 billion.

This decisive stance is supported by the materializing path of commercialization returns. Beyond AI revenue exceeding 30%, the BaiLian platform's client base grew eightfold YoY, with AI model and application service ARR targeting RMB 30 billion by year-end. Alibaba also aims to achieve full-stack self-development from GPUs and CPUs to storage and network chips, with rising self-developed chip penetration further boosting gross margins. As revenue growth and cost optimization curves both ascend, Alibaba's bold bets gain clear return expectations.

Tencent's DNA is entirely different. Its core strengths lie in social relationship chains and content ecosystems, with WeChat and QQ constructing China's largest social network, gaming maintaining global leadership, and advertising connecting millions of merchants. This DNA makes Tencent more adept at 'keeping users engaged longer in existing scenarios' rather than 'providing foundational tools for enterprise clients.'

Pony Ma has a clear understanding of this, stating that Tencent 'cannot casually invade others' territories just because they're doing well,' as 'we've tried that before and mostly failed.'

Thus, Tencent adopts a more cautious, step-by-step investment approach, refusing blind arms races. In Q1, Tencent's capital expenditures, including IT infrastructure and data center investments, reached RMB 31.9 billion (+16% YoY). Tencent's calm stems from the growth momentum of core businesses and free cash flow: domestic gaming revenue grew 6% YoY, international gaming revenue grew 13% YoY, marketing services grew 20% YoY, and free cash flow grew 20% YoY.

Tencent President Martin Lau stated during the earnings call that model training is an investment in the future, which may not yield immediate returns but will accumulate model capabilities over time, gradually unlocking numerous commercial opportunities. This 'unhurried' mindset is built on the stability of core businesses.

These two investment rhythms each have their logic and risks. Alibaba bets on the scarcity window of computing power, aiming to lock in supply advantages during demand surges and build cost moats through scale effects and full-stack self-development. Tencent bets on the technology maturity curve, waiting for model capabilities to stabilize before leveraging product experience and ecosystem advantages to counterattack.

Alibaba's risk lies in the sustained cash flow pressure from massive capital expenditures, while Tencent's risk is potentially missing the pricing power window from technological leadership. Alibaba stands at the critical juncture of translating technological dividends into commercial dividends, while Tencent uses time to gain space.

Converging Paths, Blurring Boundaries in the Ecosystem Endgame

However, Alibaba and Tencent's two paths are now extending into each other's territories following their respective logics, with boundaries rapidly blurring. This is not simple mutual imitation but an inevitable spillover as both models mature.

Alibaba grows upward from its foundation, extending infrastructure capabilities into user-centric scenarios. The QianWen App has fully integrated with Taobao, Tmall, Alipay, Gaode, and Fliggy, becoming China's first omnipotent personal assistant spanning life, work, and learning scenarios.

More significantly, Alibaba adjusted its organizational structure this March, establishing the Alibaba Token Hub business group with the core objectives of 'creating tokens, delivering tokens, and applying tokens,' elevating strategic synergy among models, MaaS, and AI applications to an independent business group level. It swiftly launched intelligent agent products like 'Wukong' and 'Miaowu.' This structural adjustment indicates that Alibaba is no longer satisfied with being a 'wholesaler' of computing power but aims to define user entry standards in the AI era.

Tencent, meanwhile, roots downward from applications, gradually addressing infrastructure shortcomings. During the earnings call, Tencent executives stated that more domestic AI chips will arrive monthly in H2 2026, meaning that while providing computing power for internal AI projects like the Hunyuan foundation model, WeChat intelligent agents, Yuanbao, WorkBuddy, and CodeBuddy, Tencent Cloud's external computing power commercialization will also gradually commence.

The reconstruction of the Hunyuan team is equally noteworthy. After appointing former OpenAI researcher Yao Shunyu as Chief AI Scientist, the team reconstructed pre-training and reinforcement learning infrastructure within months, launching the Hy3 preview model, whose Token calls exceeded the previous model's by 10x, topping the OpenRouter platform for multiple consecutive weeks.

This speed demonstrates that Tencent is not abandoning infrastructure competition but first identifying high-value scenarios, considering how to synergize products and models, and then reversely customizing foundational capabilities after validating demand at the application layer.

From a broader perspective, the blurring boundaries between the two companies are fostering a new industrial ecosystem. Alibaba provides 'utilities,' while Tencent addresses 'how to use them'—this divisional framework still holds, but the gray area between them is expanding.

As Alibaba's intelligent agent products compete for user attention and Tencent's computing power base begins serving external clients, the binary distinction between 'pure infrastructure providers' and 'scenario operators' no longer suffices. More accurately, both companies are evolving from different starting points in the AI value chain toward the same goal of an 'infrastructure + application' integrated platform.

The difference lies in Alibaba's bottom-up approach and Tencent's top-down approach. This evolutionary direction depends on who can breach the other's core barriers sooner.

Alibaba must prove that a company renowned for B-end services can also understand C-end user interaction logic—not just providing technical interfaces but creating product experiences users are willing to engage with. Tencent must demonstrate that a company built on content and product experience can also maintain patience in sustained foundational technology investments—not just following technological trends but establishing irreplaceable advantages in computing power, models, and chips.

Today, Alibaba has revealed its commercialization transparency, while Tencent continues to accumulate hidden momentum deep within scenarios. The endgame of these two paths does not hinge on who gains the upper hand in a single quarter's earnings report but on who can faster (more quickly) address their own shortcomings without being disrupted by the other's rhythm. Gigantic tankers and speedboats must ultimately sail toward the same ocean.