China’s Zhejiang province is positioning itself as a major artificial intelligence hub by expanding computing infrastructure, supporting open-source model development, and bringing AI into manufacturing and commercial applications. The strategy was highlighted at the 2026 World Internet Conference Wuzhen Summit, where provincial industry figures showed core AI industry revenue reaching 582.87 billion yuan in the first eight months of the year, a 26.9% increase year over year. The province’s approach links large-scale computing capacity with domestic AI models, developer communities, and industrial deployments, reflecting China’s broader effort to turn AI investment into measurable productivity gains.
Computing Infrastructure Becomes a Strategic Priority
Computing capacity is central to Zhejiang’s strategy because AI applications increasingly require access to large-scale infrastructure for model training, inference, image generation, and industrial workloads.
According to the report, the province had 197 EFLOPS of total computing power at the end of September, placing it among China’s highest-ranking regions. Fifty-four enterprises and institutions also received national computing-power vouchers worth a combined 408 million yuan, representing 37% of the national total.
China Mobile’s Zhejiang branch has developed an integrated computing network to connect these resources with industrial users. The report identifies the Hangzhou and Jiashan intelligent computing centers as important infrastructure additions, with the stated objective of enabling millisecond-level computing access, coordinated computing-network scheduling, and integrated security services.
The economic rationale is visible in local applications. A sock manufacturer in Zhuji uses AI to generate 20 original designs in three minutes, while traders in Yiwu use computing services to create high-resolution product images and promotional videos. These examples illustrate how computing infrastructure can support businesses that may not have the resources to develop their own AI systems.
The province’s broader digital economy expansion provides additional context. Zhejiang’s core digital economy industries generated more than 1 trillion yuan in value added during the first three quarters of 2026, up 12.1% year over year.
Open-Source Models and Developer Ecosystems
Zhejiang is also placing open-source AI development at the center of its innovation strategy. The provincial government has identified building a high-quality open-source ecosystem as a priority, alongside developing domestic large models and AI applications.
Hangzhou’s ModelScope community is one of the key components of this effort. The China Daily report says the platform connects more than 29 million AI developers and hosts over 200,000 models. Such communities can reduce barriers to experimentation by giving developers access to models, tools, and shared resources rather than requiring every organization to build foundational capabilities independently.
The presence of models such as Alibaba’s Qwen and DeepSeek also reflects the growing role of China’s domestic model ecosystem. Their continued development provides a foundation for companies and research teams building applications across manufacturing, commerce, and other sectors.
One example highlighted in the report is MoWorld, a real-time interactive world model developed by Moxin Technology. The company was founded by Chen Tianrun, a Zhejiang University doctoral student, and the model is described as running entirely on domestically developed neural processing units. This illustrates the province’s interest in connecting AI software development with domestic semiconductor and computing capabilities.
However, the scale of a developer community or the number of available models does not by itself establish model quality or commercial adoption. The longer-term test will be whether open-source access translates into useful applications, sustainable businesses, and research advances.
Embodied AI Brings the Strategy Into Physical Environments
The Wuzhen Summit also showcased how Zhejiang’s AI ambitions extend beyond software and cloud infrastructure. More than 30 embodied-intelligence robots from companies including AgiBot, UBTech, Unitree, and Zhongke Fifth Era appeared at the Light of Internet Expo.
A humanoid robot capable of playing badminton autonomously with human participants attracted particular attention. The report also describes table-tennis capabilities, highlighting the growing use of physical interaction as a demonstration of robotic perception, movement, and response.
These demonstrations connect to a wider industrial opportunity. Robots operating in factories and other physical environments require a combination of AI models, sensors, computing hardware, control systems, and reliable connectivity. Zhejiang’s existing manufacturing base offers potential deployment environments for these technologies, although exhibition demonstrations should not be treated as proof of widespread industrial readiness.
The province is also seeking to expand its innovation pipeline. The Global Internet Competition of “Straight to Wuzhen” attracted 1,347 projects and one-person company entries from 31 countries and regions after registration opened in June. Meanwhile, the Digital Economy Industrial Cooperation Conference has recorded annual project signings exceeding 100 billion yuan over the past three years, according to the report.
From AI Investment to Industrial Productivity
Zhejiang’s development strategy brings together several layers of the AI economy: computing infrastructure, semiconductor capabilities, large models, open-source development, robotics, and industrial applications. Rather than focusing only on model development, the province is attempting to connect these components with manufacturing and commercial demand.
The scale of the existing industry provides a significant starting point. Zhejiang reported 713.1 billion yuan in core AI industry revenue for 2025 and has set a target of reaching 1.2 trillion yuan in annual revenue by 2030.
Whether the province achieves that ambition will depend on more than infrastructure investment. Computing resources must be economically accessible, AI systems must deliver measurable improvements in industrial operations, and open-source development must generate solutions that companies can deploy and maintain.
The Wuzhen Summit offered a snapshot of that approach. Zhejiang is building an AI ecosystem in which computing power, domestic models, developers, and physical applications are intended to reinforce one another. Its progress will be measured not only by the scale of its infrastructure but by how effectively those resources translate into productivity, commercial adoption, and technological innovation.
4. Market Landscape
Zhejiang’s strategy reflects several converging trends in the global AI industry:
- AI infrastructure expansion: Regions are investing in data centers, intelligent computing facilities, and networks to support growing training and inference demand.
- Open-source model development: Shared models and developer communities are lowering barriers to experimentation and application development.
- Industrial AI adoption: Manufacturers are using AI for design, quality inspection, maintenance, production planning, and workflow automation.
- Domestic AI hardware: Locally developed NPUs and other accelerators are becoming part of efforts to build domestic computing supply chains.
- Embodied AI and robotics: AI models are increasingly being connected to physical systems that perceive, move, and interact with real-world environments.
Zhejiang’s competitive proposition is its attempt to connect these layers within one regional industrial ecosystem. The province already has a broad digital economy and manufacturing base, while its 2030 target of 1.2 trillion yuan in core AI industry revenue provides a measurable policy ambition.
The challenge is converting infrastructure capacity and policy support into commercially sustainable AI applications. Computing capacity alone does not guarantee utilization, and demonstrations do not establish that systems are ready for large-scale deployment. Industrial adoption, developer activity, and measurable productivity improvements will be important indicators of whether the strategy succeeds.
Top Insights
- Zhejiang reported 582.87 billion yuan in core AI industry revenue during the first eight months of 2026, up 26.9% year over year.
- The province’s computing capacity reached 197 EFLOPS by the end of September, supported by a network developed by China Mobile’s Zhejiang branch.
- ModelScope reportedly connects more than 29 million developers and hosts over 200,000 models, supporting Zhejiang’s open-source AI ambitions.
- Industrial examples include AI-generated textile designs, automated product content creation, and applications showcased by embodied-intelligence robotics companies.
- Zhejiang aims to grow annual core AI industry revenue to 1.2 trillion yuan by 2030.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
