China’s OpenClaw frenzy has moved past viral curiosity and into commerce. A cottage industry of paid bootcamps, gated WeChat communities, and freelance installation services has sprung up around the open-source AI agent framework, according to a detailed report from MIT Technology Review.
The services range from ¥3,000 to ¥10,000 ($410–$1,370 USD) for enterprise-grade OpenClaw deployment — a remarkable price tag for setting up free, open-source software. WeChat groups dedicated to OpenClaw tips and troubleshooting are charging monthly membership fees. Tutorial sellers on Xiaohongshu and Bilibili are packaging beginner guides behind paywalls.
The Cloud Giants Are Playing Along
The gig economy hasn’t gone unnoticed by China’s major cloud providers. Alibaba Cloud, Baidu, and ByteDance have all pivoted their API marketing to feature OpenClaw compatibility prominently. MiniMax and Qwen — two of China’s most aggressive domestic LLM providers — are positioning themselves as the preferred backend for OpenClaw deployments, as Fortune reported separately.
This creates a dynamic where a Western open-source tool has inadvertently become a distribution channel for Chinese AI API services competing against each other for market share.
A Pattern That Repeats
Every technology wave generates a services layer faster than it generates direct revenue. WordPress spawned a multi-billion-dollar ecosystem of agencies, theme shops, and managed hosting providers long before most WordPress sites made money. Android’s adoption in China created an entire parallel economy of app stores, ROM modders, and device optimizers.
OpenClaw is following the same playbook at accelerated speed. The framework launched in November 2025. Four months later, there are people making a living from installing it.
What This Signals
The services economy forming around OpenClaw in China is a leading indicator for what happens when open-source AI agents reach mainstream adoption globally. The tool itself is free. The knowledge to configure it, secure it, and integrate it with specific business workflows carries real economic value — especially when the tool is complex enough that most users can’t set it up alone.
For OpenClaw’s maintainers, this presents a familiar open-source tension: widespread adoption driven partly by third parties whose quality and security practices are uncontrolled. Some of those ¥10,000 deployments may be configured by people who understand marketing better than they understand the security implications of running an AI agent with system-level access.
The gold rush is real. Whether the prospectors are building on solid ground is another question entirely.