According to a recent report by IDC, "China AI Gaming Cloud Market Insights, 2025," Baidu's Intelligent Cloud division secured a 51% market share in China's AI gaming cloud sector, with its AI gaming cloud infrastructure (AI IaaS) segment reaching a 56.4% share.

Zhang Wei, Vice President of Baidu Intelligent Cloud and General Manager of its General Technology Business Department, shared insights on the market's trajectory in a recent interview. He stated that future growth in the AI gaming cloud market will not solely depend on "purchasing more GPUs" but will be driven by new, emerging payment scenarios.

Zhang identified four key categories for these new revenue streams: model training and inference; data services including data cleaning, storage, auditing, labeling, and model fine-tuning; Agent and workflow management; and continuous operations support for games after launch.

Zhang noted that current revenue primarily comes from large gaming companies. These clients have the capacity to self-train models and require extensive computing power, cluster stability, networking, data cleaning, and training services, leading to large-scale single projects. However, he pointed out that in terms of customer numbers and future growth, mid-sized game studios and AI-native game startups are becoming critical sources of new business. These entities often leverage MaaS platforms to adjust models, integrate them with their own data for fine-tuning, or build business-level Agent workflows.

When asked about Baidu Intelligent Cloud's development strategy, Zhang emphasized a focus on building a co-creative ecosystem. He envisions gaming companies contributing vertical industry expertise and creative ideas, while Baidu provides computing power, models, Agent infrastructure, and data feedback capabilities. This collaboration aims to transform demonstrations into products that can be launched, operated, and successfully commercialized.