The OpenClaw boom in China is a gold rush for AI companies

George Zhang thought OpenClaw could make him rich, even though he didn’t really understand how the viral AI agent program worked. But he saw a video of a Chinese social media influencer demonstrating how it could be deployed to manage stock portfolios and make investment decisions independently. Zhang, who works in cross-border e-commerce in the Chinese city of Xiamen, was intrigued enough that he decided to try installing OpenClaw in late February.

Zhang is one of many people in China who have recently been swept up in the OpenClaw craze. Workshops teaching people how to use an AI agent have popped up in cities across the country, drawing crowds in the hundreds. Technology companies are racing to integrate OpenClaw into their platforms, while local governments have announced subsidies for entrepreneurs who build products using it. Late last week, photos of grandparents and great-grandparents lining up to install the software spread online.

After renting a cloud server from Tencent and purchasing a subscription to the Chinese language module Kimi, Zhang was able to start chatting with his OpenClaw agent, or his “lobster,” as many Chinese call him. Initially, Chang told me, he was impressed by the AI ​​agent as he watched it quickly create a long-form market analysis based on the latest breaking news. But after a few days, his crayfish began to decline, producing only a basic chart of market trends rather than a detailed report. He asked OpenClaw to create something similar to what he did on day one, and the agent always responded that he was “working on it” before never returning any results.

Zhang’s conclusion was that OpenClaw is not designed for people like him who don’t have any programming skills. “It will tell me that I need to configure the API port. But that’s a technical task, and not something I can do unless I have a tutorial that walks me through it step-by-step,” he says. Eventually, he gave up letting his lobster trade stocks, settling instead for asking her to aggregate AI industry news, which he used to build a social media content farm on WeChat.

This week I reached out to six OpenClaw users in China about their experiences with the agent, and a clear picture emerged of the divide between tech-savvy adopters and those who aren’t. People who master AI see OpenClaw as a game-changer in productivity, but those without a technical background feel like they were promised a miraculously powerful AI product that ultimately didn’t deliver results. But by the time the bubble burst, they had already started paying for cloud servers and LLM tokens.

The real driver of China’s OpenClaw mania is not ordinary users, but rather Chinese companies that stand to benefit financially from its widespread adoption. Big tech companies like Tencent, Alibaba, ByteDance, Minimax, Moonshot, and Z.ai have seen AI productivity FOMO as a rare opportunity to get everyday people to start paying for AI services, and they are reaping the biggest rewards from it.

“A chatbot uses a few hundred tokens per conversation; a single active OpenClaw instance can consume dozens or even hundreds of times more tokens per day,” says Bo Zhao, technology analyst and founder of the newsletter. Hello China Technology. Every new OpenClaw user is someone who pays 24/7 for LLM API calls. “That’s why Tencent engineers were setting up tables outside headquarters to help people install the software for free,” he says.

“I couldn’t understand any of it.”

Song Chucun, a college student in China, says she started having problems with OpenClaw as soon as she tried to install it. Song is a social media intern at an AI startup, but he has no programming experience, so figuring out how to run OpenClaw turned out to be difficult. She asked Doubao, ByteDance’s popular AI chatbot, to create a step-by-step tutorial for her, but that didn’t help much.

“There were pages full of code, and I couldn’t understand any of it,” she says. “I kept asking the AI ​​to generate a response for me, and then I would paste it in and run it, and it would throw up an error, so I would try a new response.” Installation ended up being the most frustrating part of the OpenClaw for Song experience, and she didn’t feel like she learned anything from it.