AI competition is no longer about single-point technology competition, but rather about full-chain capability competition
Recently, Tencent's investment department has maintained an exceptionally busy schedule. In mid-March alone, Tencent participated in Zhipu AI's Series C funding round, followed by reports of investment in Baichuan Intelligence. Simultaneously, Tencent has been in intensive discussions with multiple top-tier domestic AI startups including Moonshot AI, Light Years Beyond, and MiniMax.
Whether through large language model releases, AI application deployment, computational power competition with the United States, or talent acquisition, the tech giants of the mobile internet era continue to converge in the AI arena.
01 Tencent: The Master of Balance in the AI Ecosystem
Tencent has always excelled at maximizing value through ecosystem integration.
Tencent's return to the AI table stands in stark contrast to its relatively reserved investment posture over the past two years. During 2022-2023, Tencent focused more on overseas games and B2B SaaS enterprise investments, while remaining relatively cautious in the AI sector. However, entering 2024, Tencent has significantly accelerated its investment pace in the domestic AI field.
"It's neither possible nor necessary for us to develop every AI technology direction ourselves, but we can become the greatest enabler and beneficiary of AI innovation," a Tencent investment department executive once stated during an internal meeting. This statement perhaps provides the key to understanding Tencent's AI investment strategy.
Unlike Alibaba and Baidu, which primarily pursue self-developed technology paths, Tencent has chosen to advance its self-developed Hunyuan large language model while simultaneously investing in domestic leading large model companies such as Zhipu AI and Baichuan Intelligence, avoiding the risks associated with a single technological pathway.
According to SuperCLUE's "Chinese Large Model Benchmark Evaluation Report of August 2024," in an evaluation of 11 capability items, Tencent's Hunyuan ranked first domestically in 8 core tasks. Currently, nearly 700 business scenarios (70%) within Tencent have integrated the Hunyuan model.
Tencent has long possessed enormous natural advantages in its application ecosystem. Its rich and highly adhesive application scenarios not only provide vast data resources but also serve as fertile ground for rapid AI technology deployment and scaled applications.
Through investment rather than acquisition, Tencent can establish close connections with top AI teams while maintaining their entrepreneurial vitality and independence.
Tencent's investment directions primarily revolve around core scenarios including gaming AI, content creation, enterprise services, and financial technology. Each invested company has its own characteristics: Zhipu AI has a deep academic background, with its ChatGLM series models having extensive influence in the open-source community; Baichuan Intelligence demonstrates outstanding performance in commercial applications; while MiniMax possesses unique advantages in multimodal interactions and user experience.
02 Tencent and Alibaba: Converging Paths
Behind AI investments lie three barriers: technology, application scenarios, and talent. Internet giants entering the AI arena are not merely playing a capital game; they are also building strategic moats.
In the current rapidly evolving AI technology landscape, relying solely on self-development may not allow for quick responses to the pace of technological change.
"AI competition is no longer about single-point technology competition, but rather about full-chain capability competition," according to a Tencent investment department insider. "Tencent's strategy is not to create the strongest single-point technology, but to build the most complete AI capability matrix."
Tencent's strategy can be summarized as "distributed technology integration." This model tends to bind multiple promising entrepreneurial teams through capital investment in the early stages of high technological uncertainty, building a distributed technology matrix, then coordinating integration with its own business systems through large models. The investment logic is a continuation of this concept—diversifying deployment during the initial stages of technological transformation, then proceeding with integration and deepening once the track gradually becomes clear.
The alternative is the "vertical closed-loop platform" model. Alibaba resembles more of an "AI factory" builder—controlling all key modules as much as possible, from algorithm research and development, training data, computing power platforms, and API services to end products, forming a vertical closed loop that emphasizes consistency, stability, and controllability.
Taking the Tongyi large model as an example, Alibaba has announced plans to invest over 300 billion yuan in AI infrastructure over the next three years, focusing on building a product chain around Alibaba Cloud, including DingTalk, Tmall Genie, and others. This model offers high execution efficiency and strong consistency, but its disadvantages are also very evident—once a path choice proves erroneous, adjustment becomes extremely difficult.
Comparing the two strategic models, each has its advantages and disadvantages. Distributed strategy excels in flexibility, risk diversification, and diverse innovation sources, allowing multi-directional exploration when technological paths remain unclear, avoiding betting on the wrong horse. However, its drawbacks are also obvious, namely greater coordination difficulties and complex technology path integration, potentially leading to compatibility and consistency issues between systems.
The vertical strategy, meanwhile, offers high execution efficiency, strong product consistency, and deep technological accumulation, capable of forming a complete closed loop from basic research and development to application deployment. Its disadvantage lies in potentially limited innovation vitality and high risk of single-point technology paths—once the technology direction deviates, adjustment costs become extremely high.
"Neither model has absolute advantages or disadvantages; the key is who can more quickly transform technology into commercial value," a senior AI investor stated in an interview with EqualOcean. "Tencent's advantage lies in rich scenarios and a massive user base, enabling rapid validation of AI application effectiveness; while Alibaba's advantage lies in deep technological accumulation and strong industrial empowerment capabilities, providing more systematic solutions for enterprise clients."
03 Understanding Three AI Keys: Speed, Scale, and Monetization
If Tencent's AI investment resembles a systematic engineering project rather than a "sprint," the "moat" it's constructing revolves around three main lines: technology iteration speed, application scenario scale, and commercial monetization capability.
First is technology iteration speed. In the rapidly changing AI field, whoever can iterate technology faster gains a competitive advantage.
Technology iteration includes not only model performance improvement but also algorithm optimization, training efficiency enhancement, and inference speed acceleration across multiple dimensions.
Investing in multiple AI companies disperses technology path risk, ensuring that Tencent won't fall too far behind on any technological route. Meanwhile, full-stack self-development ensures consistency and controllability of technology iteration. Although betting on a single technology path carries higher risk, once the direction proves correct, it enables rapid iteration across the entire chain from foundation to application.
Second is application scale. AI technology can only form positive feedback loops in large-scale applications—more users bring more data, more data trains better models, and better models attract more users.
For instance, Tencent, relying on its massive user base in social and content ecosystems, can promote AI applications to hundreds of millions of users in an extremely short time, creating scale effects. Product lines such as WeChat, QQ, Tencent Video, and Tencent Games provide uniquely advantageous deployment scenarios for AI applications. Alibaba, on the other hand, relies on e-commerce and cloud computing platforms to deliver AI capabilities to enterprise clients. Although its user volume doesn't match Tencent's consumer products, the value per customer is higher, and application scenarios address more rigid demands.
Lastly, and most crucially, is commercial monetization capability. Technology and applications must ultimately transform into commercial value to support continued investment. Currently, Tencent has begun to commercialize AI in areas such as advertising and gaming.
For example, Tencent's AI advertising system 3.0 has been successfully deployed, significantly improving advertising placement efficiency and conversion rates. In gaming, AI technology application has also substantially shortened content production cycles and reduced development costs. Alibaba, meanwhile, explores AI value transformation in areas such as e-commerce recommendations and enterprise services, realizing AI technology value monetization by enhancing user shopping experiences and improving merchant operational efficiency.
Currently, the capital market maintains an optimistic outlook on Tencent's AI strategy.
According to predictions from Soochow Securities Research Institute, Tencent's AI layout will create a multiplier effect on its future growth, with adjusted net profit expected to reach 265.3 billion yuan in 2026, corresponding to a PE ratio of 14 times. The dual-drive model of "self-development + investment" is expected to release enormous value in the coming years.
03 Is the Future Already Here?
From the AI industry landscape perspective, we can foresee several evident change trends.
First, the trend of AI startups "picking sides" will become increasingly apparent. As giants like Tencent and Alibaba increase their investment intensity in the AI field, high-quality AI startups will accelerate their alignment with specific ecosystems.
This trend has already begun to emerge in the market—leading large model companies such as Zhipu AI and Baichuan Intelligence have received strategic investment from Tencent, while other AI startups have chosen to cooperate with giants like Alibaba and Baidu.
This "side-picking" phenomenon will place greater survival pressure on independently developing AI companies, thereby accelerating industry consolidation.
In the coming years, we may witness further clarification of AI startup camps, with boundaries between major ecosystems becoming more distinct.
Second, "technology + scenarios" will become the core competitive key in the AI era. Pure technological innovation or scenario advantages alone can hardly build lasting competitiveness. Only by deeply combining advanced technology with rich scenarios can a true value closed loop be formed.
Tencent's investment model precisely strengthens this point, connecting technology with scenarios through capital bonds to create synergistic effects. AI startups invested in by Tencent can access massive application scenarios and data resources within the Tencent ecosystem, accelerating technology validation and commercialization. Meanwhile, Tencent can enhance its own product experiences through these innovative technologies, creating a win-win situation.
Third, existing mobile internet business lines will be reconstructed by AI. AI technology will not only improve operational efficiency of existing businesses but also catalyze entirely new business models and growth points.
Tencent has identified AI capabilities as one of its core strategic directions for the next 3-5 years and will further increase its investment in the AI field. "Tencent's advantage lies not in creating the most advanced single-point technology, but in deeply integrating technology with scenarios."
The value of technology must ultimately be realized through application, and evidence shows that Tencent's greatest advantage lies precisely in its rich application scenarios and massive user base.
The "battle of a hundred models" also proves that purely technological pursuit may not be effective; becoming a value creation point for technology integration and commercial application is the key to breaking through barriers.
We closely monitor Tencent's AI investment trends because they influence not only capital flows and talent distribution but also the entire industrial ecosystem. Whether China's AI industry can stand out in fierce global competition largely depends on the strategic choices and execution capabilities of tech giants like Tencent.
Tencent's model of investing in AI startups also provides a sustainable development path for China's AI industry—large enterprises and startups mutually empowering each other, jointly promoting technological innovation and commercial implementation.
Regardless of the final outcome, one thing is certain: the AI revolution has just begun, and Tencent has already made its opening move in this revolution.
In future AI competition, whether Tencent can transform its AI ecosystem advantages and investment layout into lasting competitiveness will determine whether it can continue to widen the gap with other internet giants.
