China’s three largest technology companies are no longer just adopting OpenClaw. They are building parallel commercial ecosystems on top of it, each betting that the open-source AI agent framework will become the foundation of enterprise automation in the world’s second-largest economy.
In the first week of April alone, Tencent launched an enterprise management platform called ClawPro, ByteDance’s Volcengine became an official sponsor and infrastructure provider for OpenClaw’s Chinese skill marketplace, and Alibaba’s Wukong agent platform expanded across its DingTalk collaboration tool. The three companies are pursuing different strategic playbooks, but the goal is the same: control the commercial layer that sits between OpenClaw’s open-source core and the enterprises that want to deploy it.
The stakes explain the urgency. China now has more OpenClaw users than any other country, roughly double the activity of the United States, according to SecurityScorecard analysis cited by The Next Web. As of late March, OpenClaw had 335,000 GitHub stars, 27 million monthly visitors, 2 million active users, and more than 13,700 community-built skills on its ClawHub marketplace. Chinese developers colloquially call setting up an OpenClaw instance “raising a lobster,” after the framework’s crustacean logo.
Tencent: The Super-App Play
Tencent’s approach is the broadest and most commercially aggressive. The company has built an entire product suite around OpenClaw, spanning consumers, developers, and enterprises.
At the consumer level, Tencent released QClaw, a mini-program that embeds OpenClaw inside WeChat, giving the framework potential access to the messaging app’s 1.3 billion users. Alongside it came ClawBot, a WeChat plugin supporting multi-modal interactions, and WorkBuddy, a workplace AI agent that according to The Next Web was tested by more than 2,000 non-technical employees across human resources, administration, and operations.
The enterprise tier arrived in early April. ClawPro, launched in public beta by Tencent’s cloud division, allows businesses to deploy OpenClaw-based agents in as little as 10 minutes. It includes controls for template selection, model switching, token-consumption tracking, and security compliance. During its internal beta, ClawPro was adopted by more than 200 organizations across finance, government, and manufacturing, according to The Next Web.
On the platform integration side, Tencent’s QQ messaging app became the first Chinese social media platform natively integrated into OpenClaw’s official codebase. The QQBot source code was merged into OpenClaw’s main repository, according to the South China Morning Post, enabling multi-account setups, slash commands, and automated reminders directly within QQ private chats.
The pattern is clear: Tencent is treating OpenClaw the way it treated mini-programs in WeChat. The strategy is to make WeChat (and QQ) the primary interface through which Chinese users interact with AI agents.
ByteDance: The Infrastructure Bet
ByteDance’s Volcengine is taking a different approach. Rather than building consumer-facing agent products, it is positioning itself as the infrastructure layer that powers OpenClaw in China.
At an event in Wuhan in late March, Volcengine unveiled plans to co-build the official Chinese mirror site for ClawHub, OpenClaw’s skill marketplace. The mirror hosts more than 43,000 skills and is designed to improve stability and access speed for Chinese users, according to TechNode. Volcengine joined OpenAI, Baidu, and Tencent on OpenClaw’s roster of official sponsors, per Caixin Global.
The infrastructure play extends to models. ByteDance’s Doubao models reached 120 trillion daily tokens in March, representing approximately 85.7% of China’s total model-as-a-service volume of 140 trillion daily tokens, according to Caixin Global. That volume doubled in three months and represents a 1,000x increase since May 2024. Volcengine also launched ArkClaw, a product integrating OpenClaw with Doubao models and ByteDance’s Feishu collaboration platform.
The strategic logic: if OpenClaw agents in China run on ByteDance’s models and access skills through ByteDance’s mirror, Volcengine captures the compute revenue regardless of which consumer-facing products win.
Alibaba: The Enterprise Workflow Play
Alibaba is the third entrant, and its bet is on enterprise workflow integration. In mid-March, the company launched Wukong, a platform that connects AI agents to handle complex business tasks including research, editing, and documentation.
Wukong is available as a standalone desktop application or through DingTalk, Alibaba’s collaboration platform that serves more than 20 million corporate users, according to Reuters. Integration with Slack and Microsoft Teams is on the roadmap. The platform will also connect to Alibaba’s e-commerce ecosystem, including Taobao and Alipay, CNBC reported.
The Wukong launch coincided with an internal reorganization that placed the AI agent platform under Alibaba’s new Token Hub business group, signaling that agent infrastructure is now a top-level strategic priority alongside cloud computing and e-commerce.
The Ground Game
The competition extends beyond software. Chinese tech companies have been running physical adoption campaigns that have no Western equivalent.
Tencent organized public “OpenClaw one-stop service day” events in cities including Shenzhen and Hefei, according to the South China Morning Post, drawing crowds of retirees and students. Baidu held similar events in Beijing. A cottage industry emerged around the frenzy: technicians began charging 500 yuan (approximately $72) for on-site installations, according to The Next Web.
Government support has amplified the momentum. “Claw-powered” one-person companies became a talking point at the National People’s Congress, and local governments began offering grants to startups building applications on OpenClaw, per The Next Web. The city of Shenzhen’s Longgang district has offered subsidies to OpenClaw-related startups, according to Global Voices.
The Security Friction
The enthusiasm has not been unchecked. China’s National Computer Emergency Response Team (CNCERT) warned that OpenClaw had “extremely weak default security configuration,” and the Ministry of Industry and Information Technology published formal security guidelines urging minimum permissions and limited internet exposure, according to The Next Web. State-owned enterprises and government agencies received notices warning against installing OpenClaw on office devices.
This security friction is, paradoxically, part of what makes the enterprise platform play valuable. The open-source version of OpenClaw was never designed for enterprise-grade data governance. Tencent’s ClawPro, ByteDance’s managed infrastructure, and Alibaba’s DingTalk-integrated Wukong all promise to solve that problem. The security gap becomes the commercial opportunity.
Three Bets, One Framework
Each company is making a structurally different wager:
Tencent is betting on distribution. With WeChat’s 1.3 billion users and QQ’s native OpenClaw integration, it controls the surface area through which most Chinese consumers will encounter AI agents. ClawPro extends that reach into enterprises that need managed deployment.
ByteDance is betting on infrastructure. With Doubao processing 85.7% of China’s model-as-a-service tokens and Volcengine hosting the official ClawHub mirror, it captures compute revenue from every OpenClaw deployment that touches its stack, regardless of which front-end wins.
Alibaba is betting on workflow integration. With 20 million DingTalk corporate users and planned integrations into Taobao and Alipay, Wukong connects AI agents directly to the commerce and payments infrastructure where enterprise value gets created.
The parallel is not Android, where one company built the platform and carriers distributed it. It is closer to the early cloud wars, where AWS, Azure, and Google Cloud each built differentiated managed services on top of the same Linux kernel. OpenClaw is the kernel. The commercial layer is the prize.
For builders outside China watching this unfold, the signal is straightforward. The world’s most competitive technology market, with the largest concentration of OpenClaw users on Earth, has decided that open-source AI agents are infrastructure worth fighting over. The question is no longer whether agentic AI will be commercialized at scale. It is which layer of the stack captures the margin.