In its fourth round of financing within half a year, the valuation jumped from CNY 10 billion to CNY 15 billion in just three months.
On October 13, Astribot announced the completion of a new round of financing totaling CNY 1 billion, led by Toukong Donghai, with follow-on investments from institutions such as Hongtai Fund and Guanghua Yangguang, and continued investment from existing shareholder Xianmu. Combined with the Series B financing completed in June earlier this year, the company, which was only established at the end of 2022, has already gathered Tencent-affiliated entities, Ant Group, ByteDance-affiliated entities, Zhongke Chuangda, as well as state-owned capital from Shenzhen, Wuxi, and other regions in its shareholder list.
1. Cable-Driven: Achieving Mass Production on a More Difficult Path
The most distinct technical label of Astribot is cable-driven technology. Using cables to replace traditional reducers for power transmission, mimicking the working mechanism of human tendons, it is theoretically lighter, more compliant, and safer. But theory is one thing; the engineering difficulty of cable-driven systems is extremely high: cables wear out, tension control is hard to stabilize, and long-life maintenance is almost unsolvable. This is also why 1X Technologies, which follows the same cable-driven route, has not yet delivered its first mass-produced robot, NEO, despite its valuation exceeding USD 6 billion.
Astribot has solved the problem of "whether it can be built" in advance. Self-developed compensation algorithms, tendon cable materials, tension control, and batch consistency verification—these engineering accumulations that may not sound sexy enough constitute its true moat—it is the world's first company to achieve mass production of cable-driven AI robots, with production capacity reaching the 10,000-unit level.
Lai Jie's judgment on this is very direct: "In my view, embodied AI is a very easy industry to make money in." This confidence stems precisely from the fact that he tackled the most difficult engineering problems first.
2. CNY 89,900: The Business Logic Forced by Pricing
In May this year, the company launched the T1 robot with a starting price of CNY 89,900. Behind this pricing is the realization of a complete set of cost control capabilities: the inherent lightweight advantages of the cable-driven structure, the self-research and self-production of core components, the supply chain bargaining power brought by the 10,000-unit production capacity, and the "Design for AI" approach that targets mass production from the design stage.
According to the disclosed order structure, orders for thousands of units in industrial and commercial services have been secured, AI retail solutions are operating at scale in six cities nationwide, and robots can independently sell products in the Palace Museum cultural and creative scenarios. The common characteristics of these scenarios are relatively fixed environments and highly repetitive movements, which are exactly where cable-driven manipulative robots can best leverage their advantages.
3. Wheeled Manipulative Route: Accumulating Data
Since its establishment, Astribot has chosen the wheeled manipulative route, without having the robots dance or box. Lai Jie's logic is: the growth of AI generalization capabilities is more reflected in manipulation abilities rather than mobility.
This choice was not very clever in the early stages. The financing pace was slower than peers, and the exposure was not as high as companies good at "creating hype". But the result of persisting in making manipulative robots is that it has accumulated a large amount of high-value physical interaction data in real scenarios. Practicing cooking and making tea, sorting items, and fall detection in nursing homes; completing sales, guidance, and interaction in cultural tourism scenarios; executing sorting and handling in industrial scenarios—the common characteristic of these data is: the internet cannot learn them, they can only be repeatedly obtained by running real machines.
When the industry shifts from "competing in models" to "competing in real data", the data assets accumulated by this company over the past two years begin to show their value.
4. After CNY 15 Billion, the Real Test Begins
Looking at the entire embodied AI track, its valuation of CNY 15 billion is not the highest. The valuations of companies such as Zhipingfang, Yinhe Tongyong, and Xinghaitu have already entered the range of over CNY 20 billion. But its differentiation lies in the fact that it is currently the only company in the CNY 10 billion club with cable-driven mass production as its core barrier, and also one of the few enterprises that have run through the complete chain of "technological differentiation → mass production capability → cost advantage → scenario orders".
Lai Jie once said: "In the future, 200 embodied AI companies will go public, but the overlap with the current 200 may be less than 10%." The weight of this statement is becoming more and more concrete as Astribot step by step pushes cable-driven technology from the laboratory to a production capacity of10,000 units.
This article is compiled based on publicly available network information, for reference only, and does not constitute investment advice.