[Summary]

China's AI cloud market has moved from a stage where competition is based on the performance of basic models to a stage where competition is based on how cheaply and deeply AI can be implemented into existing operations.

What investors should be looking at is not just how smart the model is.

The question is which company can most efficiently operate the value chain of "Model → Agent → SaaS → Free Cash Flow (FCF)".

As of 2026, the four companies at the center are Alibaba Cloud, Huawei Cloud, Tencent Cloud, and Baidu AI Cloud.

Alibaba is strong in Qwen and the private developer ecosystem. Huawei is strong in Ascend chips and government/state-owned enterprise demand. Tencent has WeChat economy and strong FCF. Baidu has a high AI sales ratio and a reversal option called Apollo Go.

However, every company has weaknesses.

Alibaba has a heavy capital investment burden. Huawei's private ecosystem is less open than Alibaba. Tencent's AI cloud alone has low purity. Although Baidu's AI ratio has increased, its FCF is still weak.

To conclude, Alibaba is the favorite for cloud infrastructure, Huawei is the line of defense for domestically produced AI infrastructure, Tencent is the most widely implemented, and Baidu has the potential for reversal.

China's AI cloud has moved from model competition to agent competition

In the Chinese AI market, there are many articles that only follow model names, such as Qwen, DeepSeek, ERNIE, Hunyuan, Pangu, and Doubao.

Of course, model performance is important.

However, looking at that alone is not enough when making investment decisions.

The main battleground in 2026 will be agent implementation, not model benchmarking.

If AI models are simply sold as APIs at low prices, price competition is likely to occur. In the Chinese market in particular, the commoditization of technology and price reductions are rapid.

What's important is whether you can add business agents, SaaS, cloud operations, and data infrastructure on top of it and integrate it into a company's daily operations.

Investors should view the situation as follows.

AI model provision
↓
AI agent implementation
↓
Embedded into enterprise SaaS and business systems
↓
High gross margins and recurring billing
↓
Free cash flow

AI clouds that cannot create this flow will have difficulty remaining profitable even if sales increase.

Four strong weapons and weaknesses

When comparing the Chinese AI cloud giants on the same playing field, it is first necessary to distinguish what they are trying to win.

Note that market share varies considerably depending on the survey target. Since the ranking changes for the overall cloud infrastructure, AI public cloud, and domestically produced GPU cloud, here we use Omdia's reports on the Chinese cloud infrastructure market in Q3 2025 as a guide.

CompanyMarket positionBiggest weaponBiggest weaknessPoints to check for investors
AlibabaLeading in cloud infrastructureQwen, developer ecosystem, and EC's military fundsCAPEX burden and company-wide FCF outflowCloud EBITA margin and FCF recovery
HuaweiStrong in government/state-owned enterprises/domestic alternativesAscend chips, Pangu, vertical integrationExpansion of private startup and developer ecosystemExpansion of non-government private sector share
TencentWeChat economic zone and FCF are strong1.4 billion contact points, advertising, games, paymentsPurity of AI cloud alone is lowMonetization of WeChat Agent economic zone
BaiduRapid progress in AI conversionHigh AI sales ratio, Apollo GoDecrease in search advertising and lack of FCFAI Cloud and Apollo Go turn FCF positive

What we want to see in this table is not just simple wins and losses.

Alibaba is a private AI cloud.

Huawei is a domestically produced AI infrastructure.

Tencent is an AI implementation of the app economy.

Baidu is a reversal from search to AI.

Even though the AI ​​cloud is the same, the route of capital recovery that they are aiming for is different.

Alibaba: king of the private sector and developers, but in an investment trough

Alibaba's AI momentum is strong.

Cloud Intelligence Group's revenue for the fourth quarter of the fiscal year ending March 2026 was 41.626 billion yuan, an increase of 38% compared to the same period last year. Sales to external customers accelerated to 40% growth.

AI-related product sales were 8.971 billion yuan. This was the 11th consecutive quarter of triple-digit growth compared to the same period last year.

Qwen is also strong.

As of March 2026, the Qwen family is reported to account for over 50% of global open source model downloads, with cumulative downloads reaching approximately 942 million times.

Alibaba's strength is that it can gather developers on Qwen and guide them to inference, learning, storage, MaaS, and agents on Alibaba Cloud.

However, it is quite heavy from a financial perspective.

Alibaba's overall operating loss for the quarter was 848 million yuan, with adjusted EBITA of 5.102 billion yuan, down 84% year-on-year, and FCF outflow of 17.3 billion yuan.

AI cloud growth looks real.

But the investment burden is also real.

Looking at Alibaba, I would like to see a recovery in Cloud Intelligence Group's EBITA margin and company-wide FCF rather than cloud sales.

Huawei: A child of national strategy, but challenges to the private ecosystem

Huawei is not a listed company.

Therefore, it is a little different to compare Alibaba, Tencent, and Baidu as investment targets by buying stocks directly.

Still, when talking about China's AI cloud market, Huawei cannot be left out.

The reason is the strength of vertical integration, including Ascend chips, Huawei Cloud, and Pangu models.

As the United States continues to restrict the export of high-performance GPUs, the importance of domestically produced alternatives is increasing for the Chinese government, state-owned enterprises, finance, communications, and public infrastructure. Huawei is at the center of this trend.

In particular, Baidu and Huawei are reported to have a large presence in the domestic GPU cloud market. There are also reports that Baidu had a 40.4% share of China's domestic GPU cloud market in the first half of 2025, and Huawei had a 30.1% share.

Huawei's strength lies in its policies and supply chain.

On the other hand, there are also weaknesses.

When it comes to penetrating private startups and the global developer community, it's hard to see an ecosystem as open as Alibaba's Qwen or Alibaba Cloud.

Strong in government-affiliated and state-owned enterprises. But whether it becomes a natural gathering place for private developers, SaaS companies, and startups is another matter.

When looking at Huawei, we want to see not only the tailwind from domestically produced alternatives, but also the expansion into the private ecosystem.

Tencent: Can we create a WeChat Agent economic zone?

Tencent is not a company that will compete head-on with Alibaba or Huawei for AI cloud sales alone.

Its strengths are its existing businesses, FCF and WeChat economic zone.

In the first quarter of 2026, Tencent's sales were 196.458 billion yuan, an increase of 9% compared to the same period last year. Non-IFRS net profit was 69.8 billion yuan, up 11% year on year. FCF was 56.7 billion yuan, an increase of 20%.

While Alibaba and Baidu are draining FCF with AI investments, Tencent continues to pour out cash.

This company's AI is more effective within WeChat than in the external cloud.

The total MAU of Weixin/WeChat is 1.432 billion. In advertising, AIM+ is responsible for approximately 30% of advertisers' Marketing Services spending, and total usage time for Video Accounts has increased by over 20% year-on-year. Business Services sales also increased 20% due to growth in cloud computing, including AI-related demand.

The focus from here on is the WeChat Agent economy.

Naturally incorporate AI agents into WeChat, which users and companies use on a daily basis. Expanding to advertising conversion, mini-shop automation, payments, customer support, business communications, and corporate SaaS.

Tencent has a strong ability to slip AI into everyday contact points, rather than the flashiness of its AI models.

The weakness is that the purity of the AI ​​cloud is low and easy to see.

However, there are times when the ability to convert funds into FCF becomes more important to investors.

Baidu: The AI Shift Is Fast, But Monetization Remains Difficult

Baidu has the fastest pace of AI shift among the four companies.

In the first quarter of 2026, Baidu General Business' online marketing sales were 12.6 billion yuan, down 22% from the same period last year. The slowdown in search advertising is quite severe.

On the other hand, Baidu Core AI-powered Business was 13.6 billion yuan, an increase of 49% year-on-year. The proportion of General Business was 52%, and AI-related sales exceeded conventional advertising.

This is a big change.

AI Cloud Infrastructure was 8.8 billion yuan, an increase of 79% year-on-year, and AI-native Marketing Services was 2.3 billion yuan, an increase of 36% year-on-year. Apollo Go provided 3.2 million fully driverless rides in the first quarter of 2026, and the cumulative number of rides exceeded 22 million as of April 2026.

However, Baidu's problem is cash.

Operating cash flow was a positive 2.670 billion yuan, but FCF after capital investment was minus 3.246 billion yuan on a Baidu Inc. basis.

There is a difference between having a high AI sales ratio and having strong FCF.

For Baidu to be seriously reevaluated, AI Cloud and Apollo Go need to contribute not only to sales but also to the company's overall FCF becoming profitable.

Comparison of 4 companies: Data that investors should look at

As of 2026, this is how the four companies are arranged from an investor's perspective.

Comparison itemsAlibabaHuaweiTencentBaidu
Main battlefieldAI cloud, Qwen, EC collaborationDomestic AI infrastructure, government/state-owned enterprisesWeChat, advertising, games, SaaSAI cloud, AI advertising, Apollo Go
The biggest weaponQwen and the developer ecosystemAscend, Pangu, vertical integrationFCF and WeChat contact pointAI sales ratio and autonomous driving
FCF/cashFCF outflow in the most recent quarterDifficult to compare details as it is unlistedContinued to be strongly positiveStill negative
WeaknessesDecline in CAPEX and company-wide profitsExpansion of private ecosystemLow purity of AI cloudDecrease in search advertising and lack of FCF
Confirmation pointsCloud EBITA and FCF recoveryPenetration into the private sector, overseas, and developersWeChat Agent economic zoneFCF contribution of Apollo Go and AI cloud

The most important thing in this comparison is not the size of sales.

This is the route to converting it into cash.

Alibaba moves from Qwen to the cloud.

Huawei to domestically produced chips and government/state-owned enterprises.

Tencent moves from WeChat to advertising, payments, and SaaS.

Baidu goes to AI search, AI cloud, and robotaxis.

Four companies are trying to turn AI into cash through completely different paths.

Common risks of China AI cloud

There are also risks common to all four companies.

Price competition

In China's AI cloud market, competition to lower API usage fees and cloud prices is likely to occur.

Even if demand increases, if prices fall, profit margins will not increase.

In particular, it is dangerous to only look at token usage and GPU utilization.

Investors should look at gross profit margin, EBITA margin, and FCF, not sales.

Computing resources and domestic chips

The US export regulations on high-performance GPUs are a major constraint for China's AI cloud.

Domestic AI chips, such as Huawei Ascend, Alibaba T-Head, and Baidu Kunlunxin, are becoming increasingly important.

However, having a chip and being able to create a developer ecosystem like Nvidia CUDA are two different things.

Here again, the focus is not just on the amount of supply, but on having a software platform that is easy for developers to use.

Regulation

Content review of generated AI, data management, advertising regulations, gaming regulations, and autonomous driving permits can all affect business performance.

Baidu's Apollo Go, in particular, has to deal with differences in permits, insurance, accident response, and regulatory differences in each city when expanding overseas.

Huawei is also strong in government-related demand, but is susceptible to the effects of geopolitics.

Conclusion: AI Sales Do Not Decide The Winner; FCF Conversion Does

The winners in China's AI cloud market will depend on what they value.

If you value the private sector and developer ecosystem, look no further than Alibaba.

If you want to focus on domestic substitution and government/state-owned enterprise demand, choose Huawei.

If you value financial stability and implementing AI into everyday apps, go with Tencent.

If you value the structural transformation from search to AI and the options of Apollo Go, choose Baidu.

However, what investors should ultimately be looking at is the same.

The decisive factor is not AI sales alone, but FCF conversion power.

Build the model.

Turn it into an agent.

Incorporate into SaaS or business systems.

Continuous billing with high profit margins.

At the end, cash remains as FCF.

The company that can create this trend fastest and at the lowest cost will be the real winner in China's AI cloud market.

This article is intended to summarize the thinking behind investment decisions, and is not intended to recommend buying or selling specific stocks. Huawei is a private company, and general investors cannot directly buy or sell its shares. Chinese stocks, Hong Kong stocks, and US-listed ADRs are subject to risks associated with price fluctuation risk, exchange rate risk, liquidity risk, regulatory risk, geopolitical risk, and differences in accounting and disclosure systems.

Related articles

Source

This article is for educational and informational purposes only, based on public information. It is not a recommendation or solicitation to buy or sell any specific security or financial product. Although care is taken with accuracy, the content and future investment outcomes are not guaranteed. Final investment decisions should be made at your own judgment and responsibility.