Tencent Holdings demonstrated significant progress in AI applications and infrastructure construction in the second quarter of 2026. The company has advanced the development of AI models, such as the Hunyuan of Swiss production versions, enhanced productivity tools and services at the application layer, such as AI office assistants and AI coding assistants, and increased the procurement of computing resources to support future revenue growth. Financial results show that total revenue, gross profit, non-International Financial Reporting Standards (IFRS) operating profit and net profit have increased to varying degrees. In addition, the number of monthly active users of social networking services such as WeChat and QQ was stable, and digital content, games and cloud services also performed well.

On the basis of the steady growth of the existing business, Ten-Point Technology continues to promote the AI strategy. By building a strong basic model system, such as the optimized "Huan Yuan Three", it improves performance and cost efficiency, accelerates model iteration, deepens product integration, promotes the research and development of "Huan Yuan Four", and aims to achieve AI-driven new business growth. The company also optimizes model training through intensive learning and multi-modal capability upgrades to ensure efficient performance in complex tasks, while exploring the flexibility of external rental infrastructure to ensure the sustainability of AI investments.

Invest in a self-built base model to optimize costs, innovate features, and enhance the value of AI in office applications. It focuses on the successful cases of office productivity tools 'Workspace' and 'Code Assistant', as well as the functional upgrading and market expansion strategies of AI products such as 'Xiaowe' and 'Yuanbao, 'emphasizing the role of AI technology in improving user experience and promoting the realization of enterprise relationships.

Tencent's total revenue increased by 11% in the second quarter, with social networks, domestic games, international games, marketing services and financial technology and corporate services contributing 16%, 23%, 9%, 21% and 30% of revenue respectively. Value-added services revenue reached 98 billion billion, up 14% year-on-year. Domestic game revenue increased by 17%, thanks to games such as "Delta Force" and "PC Variant"; although international game revenue fell by 1%, it benefited from the growth of games such as "Weathering Waves. The overall viewing time of the video number has increased by more than 20%, and the time of Mini Shops GMVis has increased.

Marketing service revenue increased 22% year-on-year, thanks to the increase in UCPM and advertising display volume, AI marketing plus code intelligent delivery, video advertising and small program marketing performance outstanding. Revenue from fintech services 60 billion, with significant growth in commercial payments, wealth management and consumer loan services. The cloud business is expanding rapidly, driven by AI demand, international expansion and rising service pricing, with a surge in GPU leasing, model services and coding usage.

In the second quarter of 2026, the company's total revenue reached 20.48 billion, up 11% year-on-year, and net profit 6.84 billion, up 9%. Gross margin increased to 58%, mainly due to increased revenue from high-margin games and marketing services. R & D investment increased by 35% to 2.72 billion for AI model optimization and AI capability development. Free cash flow was negative 1.38 billion, mainly due to increased investment in AI infrastructure.

Discussed the latest quarterly capital expenditure growth of the Group and its impact on the AI business, emphasizing the rapid recovery of costs by leasing computing resources, while investing in its own models and application development, is expected to achieve higher economic benefits in the long term, the future AI of the original business will become an important profit point.

The positioning and competitive strategy of AI platform in the market are discussed, including the market segmentation of first-party and third-party applications, and the commercialization path of AI cloud services. It emphasizes the importance of AI technology to enhance user experience and ecosystem value, and discusses the profit model adjustment of the domestic cloud market in the price-sensitive environment, as well as the opportunities and challenges brought by the large-scale application of AI technology in the future.

The model cost efficiency and differentiation strategy are discussed, as well as the priority of cloud services as a high-return investment. The practical application value and future development direction of the model are emphasized, and the immediate revenue potential and long-term economic value of cloud services are pointed out.

Discusses the future vision of the AI agent transaction cycle, including the interaction between users and merchant agents to execute transactions, as well as the challenges and advantages of in-device inference technology, emphasizing the role of AI models in the growth of marketing services and advertising market trends.

The dialogue discussed how Tencent will balance capital expenditures with share repurchase activities in the early stages of the AI investment cycle, emphasizing that capital allocation will be dynamically adjusted according to the environment, giving priority to high-return capital expenditure projects, such as increasing computing power to support model building and service delivery, while possibly reducing cash investment in share repurchase.

Discusses capital expenditure planning in AI areas, emphasizing that the initial major investment in building AI-native IT will bring long-term returns, rather than increasing investment every year. At the same time, it analyzes how to create additional business value through cutting-edge models under the premise of ensuring cost efficiency, including but not limited to improving work efficiency, enhancing service capabilities and opening up new business models. In addition, the economics of AI products is discussed, pointing out that controlling models, reasoning and computing power are the key to achieving profitability.

Discusses the process of AI product investment from caution to acceleration, dynamically adjusting investment based on product performance such as user growth and revenue potential, emphasizing long-term perspective and flexible budget allocation, and strategies to shift to profit models when necessary.

AI investment priorities are discussed, including model training and cloud services, as well as the role and investment scale of the Xiao Wei project. The low profit expectation in the initial stage of the AI project was emphasized, but the long-term return was considerable and the investment will follow strict management. Quickly generating revenue by increasing computing power is mentioned as a coping strategy.