China Tech Earnings: BABA, BIDU and TCEHY

China’s technology giants have been spending hundreds of billions of yuan on artificial intelligence (AI). Now investors want to see returns. 

Gaming and social media giant Tencent [TCEHY], search engine firm Baidu [BIDU] and e-commerce heavyweight Alibaba Group [BABA] are expanding AI infrastructure, developing frontier models and integrating AI across their business. 

But are these investments bearing fruit? This earnings review examines Tencent’s AI infrastructure optionality, Baidu’s efforts to revive its search business and Alibaba’s newfound dominance in open-source AI models.

Tencent: Core business supports AI investments

Tencent’s Hong Kong-listed shares have fallen more than 24% in 2026, as of 19 August, as investors look for signs that the company’s costly AI investments can translate into earnings growth.

Its Q2 results, reported on 12 August, showed some encouraging evidence. Tencent said its AI ad recommendation model helped marketing services revenue grow 22% year-on-year to RMB43.6bn, supported by higher marketing spending on its platform. Meanwhile, demand for cloud services and AI-powered office applications such as WorkBuddy and CodeBuddy helped fintech and business services increase 9% y/y to RMB60.3bn.

Domestic gaming revenue was a bright spot, rising 17% y/y to RMB47.3bn in Q2, on the back of the soaring popularity of games such as Delta Force and Valorant in China. 

For Tencent, which is playing catch up to nimble startups such as Moonshot, DeepSeek and MiniMax in developing frontier AI models, its reliable income from legacy businesses is providing much-needed support as AI investment continues to soar.

In Q2, Tencent reported negative free cash flow of RMB13.8bn as capital expenditures surged 176% y/y to RMB52.8bn, most of which was spent on AI infrastructure.

President Martin Lau said during the earnings call that Tencent was comfortable in making significant investments in AI due to the “substantial upside potential” and “clear downside protection” referring to the company’s openness to renting out excess computing capacity “if needed.”

Overall, the Q2 results showcased the strength of Tencent’s core businesses but highlighted the rising cost of its AI buildout, as its R&D expenses surged, capital expenditures remained elevated, cash flows turned negative and profit grew just 3% from a year ago.

Key developments to watch for include Hunyuan AI model updates, WeChat AI adoption, long-term agreements for compute sales and game releases.

Baidu: Racing against an existential threat to traditional internet search

Baidu, often referred to as “China’s Google,” saw its revenue fall for the fifth consecutive quarter in Q2 as its traditional internet search and advertising business continued its decline in the AI era.

On 18 August, Baidu’s Nasdaq-listed shares dropped 12.7% to a near one-year low after its Q2 results fell short of Wall Street’s revenue and earnings expectations.

Quarterly revenue slipped 4% y/y to RMB31.3bn. Traditional online marketing services revenue was the biggest drag, slumping 19% y/y to about RMB13.1bn. The segment accounted for nearly 42% of total quarterly revenue, down from about 49.5% a year ago.

Net income attributable to Baidu fell over 68% y/y to RMB2.32bn in Q2, largely due to lower fair-value gains on long-term investments compared with the year-ago period.

In a sign of where Baidu’s business is headed, its AI-powered operations increased their share of total revenue to nearly 40% in Q2 from about 30.5% a year ago.

Baidu generates AI revenue from cloud infrastructure, AI applications and AI-native marketing services. AI cloud infrastructure was the main growth driver in Q2, with revenue rising 50% y/y. 

AI applications revenue, however, grew just 3%, while AI-native marketing services growth was flat in Q2. This has left investors cautious about Baidu’s ability to monetise AI, particularly as the company bets on a full-stack AI approach and looks to AI-powered search and recommendation systems to revive its flagship search engine.

Against this backdrop, Baidu continues to face growing pressure to restore its once-leading ERNIE large language model to the AI frontier. CEO Robin Li said that going forward ERNIE’s capabilities will be developed to focus on what matters most to Baidu’s applications, as “no single model can lead in every dimension at all times”.

“Take AI search as an example. When we improve ERNIE’s ability to understand user intent and assess content quality, we apply those improvements directly to search and feed,” Li explained.

Capital expenditure nearly tripled y/y to RMB11.39bn in Q2 as Baidu continued to expand its data centre infrastructure to train its own models and provide computing power to cloud customers. As a result, Baidu reported negative free cash flow of RMB7.95bn, an increase from negative RMB4.68bn reported a year ago.

Key developments to watch for Baidu include ERNIE updates, robotaxi business Apollo Go’s global expansion, proposed spin-off of chipmaking unit Kunlunxin and its listing in Hong Kong, transition to a dual primary listed status in Hong Kong and potential inclusion in the Hong Kong Stock Connect programme.

Alibaba: Qwen’s popularity provides a leg up in AI race

Alibaba Group Holding entered the August earnings season with momentum, extending its rally from a June-low of $91.99 to $128.90, as of the 19 August close, marking a return of over 40% for the period.

The company has gained favour among tech investors as its Qwen family of AI models has soared in popularity among developers and users. In early August, Alibaba released its latest AI model called Qwen 3.8-Max, which the company claimed performs on par with leading frontier models such as Anthropic’s Fable 5.

A 15 August Bloomberg report called Qwen “the world’s No. 1 AI model” having accumulated more than 3bn downloads from January to early August.

A report by open-source AI hub Hugging Face added that regular updates, wide use cases, and flexibility around customisations, redistribution and commercial use have helped Qwen become “part of the default workflow for developers”.

“A model’s ecosystem position is not defined only by its own releases, but by how much the community builds on top of it,” noted Adina Yakefu, Irene Solaiman and Apolinário for Hugging Face.

Alibaba is now looking to monetise Qwen’s popularity by seeking a share of the revenue generated by companies that redistribute Qwen as a service or integrate it into their applications.

Growing AI-related revenue beyond its booming cloud infrastructure business will be a key focus for investors awaiting Alibaba’s Q1 results for the period ended June 2026, scheduled for 20 August.

In its Q4 results for the period ended March 2026, Alibaba missed both revenue and earnings expectations. Its stock still rose more than 8% post-earnings on 13 May as management reassured investors that ongoing investments will help the company “firmly cement our absolute market leadership position” in the AI race.

Investors will be looking for further evidence of that progress in the upcoming Q1 update after Alibaba reported a 38% y/y surge in cloud revenue to about RMB41.62bn in Q4.

Total revenue grew 3% y/y to RMB243.38bn in Q4, with growth weighed down by the disposal of hypermarket operator Sun Art and department store chain Intime.

Free cash flow turned negative in Q4 at RMB17.30bn compared to an inflow of RMB3.74bn in the same quarter a year ago, highlighting the cost of Alibaba’s aggressive investment push.

Alibaba’s New York-listed ADRs were down 16.47% year-to-date, as of 19 August.

Key developments to watch for Alibaba include continued AI cloud revenue growth, Qwen monetisation, chipmaking unit T-Head growth strategy and capital expenditure outlook.

Source: Yahoo Finance

Conclusion

To summarise, the strength of Tencent’s core gaming and advertisement business, alongside the network effects coming from its widely popular super-app, provides the foundation to pursue its AI ambitions. 

Similarly, Alibaba’s e-commerce dominance and Qwen open model popularity buttress an already booming AI cloud infrastructure business. 

Meanwhile, Baidu’s full-stack AI strategy spanning cloud infrastructure, frontier models, enterprise applications and chipmaking combines with its robotaxi business to give it a distinct position in China’s AI landscape.