en.Wedoany.com Reported - On July 6, Frost & Sullivan released the "2025 China Full-Stack AI Cloud Services Market Report." According to the report's statistical scope, the total market size of China's IaaS, PaaS, and MaaS in 2025 reached RMB 59.59 billion, with Alibaba Cloud's total revenue hitting RMB 23.9 billion, securing the top spot with a 40.1% market share.
The report breaks down AI cloud services into three layers: the infrastructure layer, the platform layer, and the model service layer. The key change is that competition among cloud computing vendors is no longer solely about server, storage, and basic computing power scale. With the deployment of large model training, inference, agent development, and industry applications, customers require a full suite of services ranging from computing power scheduling, model invocation, and development tools to application deployment. Alibaba Cloud's 40.1% share indicates that its advantage extends beyond traditional cloud resource scale to include model services, inference platforms, developer tools, and enterprise AI application capabilities. Baidu Cloud, Volcano Engine, and SenseTime ranked second to fourth with shares of 16.6%, 13.4%, and 7.1%, respectively. The combined share of these three companies is 37.1%, lower than Alibaba Cloud's individual share. This gap suggests a clear trend of head concentration in China's AI cloud services market, where leading vendors are packaging models, platforms, computing power, and application ecosystems into more comprehensive service capabilities.
The measurement methods for AI cloud services are also evolving. In the past, enterprises focused more on GPU count, server configuration, and computing power rental duration. Now, an increasing number of tasks are shifting toward token invocation, inference throughput, context caching, model response speed, and agent execution efficiency.
For the industry chain, the expansion of the full-stack AI cloud services market will directly drive demand for AI servers, data center networks, optical modules, liquid cooling systems, power systems, storage devices, databases, middleware, model platforms, and security services. When deploying AI applications, enterprise customers not only procure a batch of computing resources but also consider model training, model fine-tuning, inference costs, permission management, data compliance, agent tool invocation, and business system integration. Competition among companies like Alibaba Cloud, Baidu Cloud, Volcano Engine, and SenseTime will increasingly revolve around model capabilities, cloud platform efficiency, industry solutions, and developer ecosystems. The market size of RMB 59.59 billion indicates that China's AI cloud services have moved from the early experimental stage into an accelerated commercialization phase. The subsequent competitive focus will shift from "who can provide computing power" to "who can stably connect computing power, models, and business scenarios."
