When a hardware company lists, everybody repeats the valuation. The number worth reading sits further down the prospectus: who is buying with their own money, and how long they cannot sell.
Unitree priced its Shanghai listing on August 6. Ten percent of the company, around 600 million dollars raised, close to six billion for the whole thing.
The Engineers Own 11 Percent of It
Before the listing, the employee shares sit in one holding company, and that vehicle owns 10.9 percent of Unitree. At the listing price that is roughly 640 million dollars belonging to staff.
It goes back to September 2017, when seventeen employees got the right to buy shares at one yuan each. That share is now worth about forty dollars, roughly a hundred times their money, and the programme ran several more times through the same vehicle.
The part that surprised me is what happened next. Staff put another 38 million dollars of their own money into the listing itself. One fund covers 161 people with 30 million dollars and a one year lock: engineers, team leads, production and R and D staff, between 140,000 and 560,000 dollars each. The second fund is ten people, 7.5 million dollars and a three year lock, almost all of them from R and D. The founder, Wang Xingxing, who already owns about a third of the company, put two million into the three year fund.
Every time I walk into a restaurant and see the owner's family eating their own food, I know it is a good restaurant.
The people closest to the product bought more of it, and they accepted the longest lock to do it.
What the Valuation Is Standing On
Last year Unitree sold 237 million dollars worth of robots, with revenue up more than tenfold in two years, and humanoids are now half of that. It shipped 5,500 humanoids in 2025, and the head of industrial engineering told me on my last visit that the target for this year is four times that. Its robot dogs hold sixty percent of the world market for consumer legged robots.
Gross margin is sixty percent, high for hardware, and the reason they gave me is that they make more than ninety percent of the components in house. They expect 150 million dollars of revenue in the first half of 2026, with 516 employees on the books.
Wang Xingxing was born in 1990 and built a four legged robot on his own during his master's, the board, the mechanics and the control software. He worked at DJI briefly before founding Unitree in 2016, and the average age on the team today is 31. Whether six billion is a fair price on 237 million of revenue is a real argument, and I do not think it is settled.
Two Hundred Million Dollars and Nothing to Sell
A second company raised 200 million dollars in a month with no product on the market and no sales at all. More than 100 million in early July, the largest single first round in Chinese consumer robotics, then another 100 million on August 3.
The founders came out of Tsinghua in 2023. The CEO was a senior scientist and product manager at DJI who ran one of its small consumer drones, and the head of AI ran its intelligent flight system. They are building a home robot. One investor explained the cheque plainly: this is the only team in China that has mass produced AI hardware by the million and can also build the models and do the engineering.
Nearly twenty companies founded by ex DJI people were funded in a single year, a five year high, and almost all of them sit within five kilometres of DJI's Shenzhen headquarters. Bambu Lab was started by the man who ran DJI's consumer drones, EcoFlow by the man who built its battery department, and Wang Xingxing passed through as well.
One investment firm rented a permanent office across from the DJI towers, not for a deal, just to watch who walks out. A department head who says he is thinking of leaving gets a three million dollar valuation before he has decided what to build.
A drone is a robot that flies, so flight control, computer vision, embedded software and precision manufacturing are the same toolkit, and DJI trained engineers who understand both the algorithm and the supply chain. When you meet a Chinese robotics company, ask whether the founder or the CTO came out of DJI.
Half a Billion Dollars for a Sense of Touch
The third company makes tactile sensors, was founded in 2021, and raised another 140 million dollars last week. That takes it to about half a billion in total, more than any touch sensor company anywhere. The round was led by a semiconductor and consumer electronics group alongside the investment arm of Bank of China and a national fund. In the past twelve months the company consumed nearly a million of its own tactile chips, which puts it well past the prototype stage.
Every humanoid you have seen can see. Cameras are cheap and vision models are good. Feeling is the part nobody has solved, and the gap shows at the worst moment: once the hand closes around an object, the camera cannot see it any more. Their sensors read fifteen things at once, including six axis force, texture and softness, down to a hundredth of a Newton, and their own humanoid carries 1,956 of them. They are also building factories whose only job is collecting touch data, one in Tianjin and three more going up.
This is the part that decides the outcome for a buyer. Two humanoids can look identical on stage and be completely different machines, and the difference is mostly in the hands. I regularly see customers arrive with a brand in mind and a problem that machine cannot actually touch.
What a Week of Funding Does Not Settle
None of this is demand. A listing price is what investors will pay, not what customers ordered, and 200 million for a team with no product prices a set of CVs. The hard numbers are the shipments, the tenfold revenue and a million tactile chips consumed, and whether deliveries keep pace with the money shows up in the first half figures, not the share price.
[00:00] You have seen this robots doing kung fu on the internet. The company that builds it is called Unitree and this month it goes public here in Shanghai where I am sitting right now. And Unitree's employees own 11% of the company. That is $600 million shared between the people who built the robots. I will tell you more today. Welcome to Asia Bits Robotics news. Our second story I want to share with you today is about a robot startup here in China that raised $200 million in 1 month. That is [music] 30 days for those who don't know. And the funny thing is it has no product and not a
[00:48] single sale yet. The only USP it has is that every founder came out of DJI, the company that makes the drones you know. Right now, worked at DJI is the most valuable line on a Chinese CV. This third story I want to bring you is also crazy. It's about a company called Paxi you. You have probably never heard of them and they just became the best [music] funded touch sensor company in the world. Half a billion dollars all of it so a robot can feel what it is holding. The founders come out of the Tokyo lab that built the first humanoid camera decades ago. [music] And then in the end, we go to Ant Group, the company
[01:37] behind Alipay. They are building a brain for robots and today in the morning I stood right next to one of their machines [music] while it was working in a pharmacy. So the robots are coming and they're already here in China. And at the end, as always, I have another cute robot video for you. So, stay tuned. Now, there's one interesting [music] thing about all the stories we want to talk about today. A lot of people in the West have never ever heard of [music] these companies, but they make a lot of money, they raise a lot of money, and sooner or later, news for you, they will come into your life in the West. That's why I do this program. That's why I produce these news. I live in Shanghai.
[02:26] I speak [music] Chinese, and I read the stuff every day. Most of the stories were only published in Chinese. So, I read them for you, and [music] I translated them. And over the past year, I have taken more than 100 VCs, board members, and founders into robotics companies here in China, in Shanghai, in Hangzhou, in Suzhou, and in [music] Shenzhen. So, if you want to walk one of these factory floors yourself, the link to contact me is in the description. Write me now, and let's get started with the Unitree story. Asia base. Unitree is a company I keep going back to. It's not far from here. It's in Hangzhou, which is around 45 minutes train ride. I've been there four, five,
[03:15] multiple times. And every time I go there, there are hundreds of people in their showroom. These people are watching their robo dogs, watching their humanoid robots do kung fu, and do the dancing. Most people have only known this company for about 3 years. I would say most of you maybe this year when you saw [music] them dancing on the kung fu on the spring festival gala. But, it was actually founded in 2016. It's 10 years old now, and they started with robo dogs. Quadruped were their first product. So, this year they go public on the Shanghai Tech Exchange and the price was set just last week on August 6th and the investors put their money in on August 10th. They're selling 10% of the
[04:03] company and raising about $600 million. That values the whole company at close to $6 billion and we're talking about US dollars here. And so, let's have a look at their numbers. Last year they sold 200 and $37 million worth of robots. Revenue is up more than 10 times in just 2 years. So, humanoids are now half of that revenue. They shipped 5,500 humanoids last year in 2025 and last time I was there, the head of industrial engineering told me that they want to do four times of that amount this year. That means at least [music] 20,000 humanoids from Unitree sold. And their
[04:51] robot dogs hold 60% of the world market for consumer legged robots. So, they already conquered 60% of the world market. The gross margin is 60% which, if you are in the hardware industry, you know it's a really, really high for a hardware company. They told me the reason themselves, almost everything they make is made in house. So, over 90% of the components they do themselves and they expect 150 million dollars of revenue for the first half 2026. And they had 516 employee employees at the end of this year. And this is not show, this is not in a computer, this is not software, this is a real manufacturer with real margins. So, now the question is, where
[05:40] does the 11% come from? And so, let me break this down for you. Let's talk about the 11%. before the listing the employees shares sit in one company that holds them together. That company owns 10.9% of Unitary this company and it's at $6 billion valuation that means about [music] $640 million belong to Unitary staff. Where does it come from this money? In September 2017 9 years ago Unitary gave 17 employees the right to buy shares at one Chinese [music] yuan each and the price now is about 40 US dollars. So roughly a hundred times [music] their money and they ran the
[06:27] same program several more times after that always through the same holding company we [music] just talked about. Here's the part that surprised even me. The staff are buying [music] even more shares in the listing itself. $38 million of their own money to buy the shares of their employer. This one runs through two funds. Fund one is a hundred and sixty-one people $30 million US dollars and [music] they cannot sell for year and who can buy these engineers, team leads, production and R&D [music] staff. They're putting in between a hundred and forty thousand and five hundred and sixty thousand dollars each in this fund and it's all [music] their own money. The finance chief, the CFO and one more
[07:16] executive put in about a million dollars each. Then let's talk about fund number two. It is only 10 people and they put seven and a half million dollars inside and they cannot sell for three years and almost all of them are from the [music] R&D department. So what does it mean? The people closest to the product are the ones buying more of it. So that is a very interesting signal. [music] You know, when I read this, I was thinking every time when I walk into a restaurant and I see the owner or the owner's family sitting there eating their own food, I know it's a good restaurant. And that is what this Unitree story reminds me of. So, now let's talk about the founder. His name is Wang Xingxing,
[08:04] which is a really nice name. It means like going up, being happy. He was born in 1990, so he's 36. I have to say sadly 2 years younger than me and much more successful than me. And he owns about a third of the company. And at a $6 billion valuation, and he put $2 million of his own money into fund number two, the one where you cannot sell for 3 years. I mean, it's maybe not a lot, but it is a signal, you know. He took 3 years and said, "I will put $2 million of my own money inside." Unitree is actually a very young company. The core engineers, they're all born in the '90s, too. And last time I was there, they told me that the average age on the team is 31. So, very young people, a very young
[08:52] team. And everyone I talk to besides the young age, they are very [music] sharp, and it's always a pleasure to talk to them. Wang Xingxing, the founder, grew up in Ningbo. He was very bad at English, so he missed the university he initially wanted to go to. And during his master's, he built [music] a four-legged robot on his own, the motorboard, the mechanics, and the control software, all of it. And after he graduated, he worked at DJI for a short time. We'll talk about DJI later. Then he left DJI and he founded Unitree in Hangzhou in 2016. And they became famous for their G1 humanoid because it sells for only $14,000. That price is what ended the idea that a
[09:41] humanoid is lab equipment just for people who have the budget. [music] Uh it put them into universities, into labs, and into factories, and soon maybe in your home. And this whole [music] IPO from Unitree and the engineers putting their own money there is a message to every engineering graduate in China. And the message is robotics is where you get rich now. The robotics industry here will get a lot of CVs in the coming years. This is also the reason why Chinese robots keep getting cheaper and keep getting better so fast. It is the engineering talent, not just the subsidies. A lot of people talk about the subsidies, but I think the talents are much more important. And what does
[10:30] it mean for you? For your business, this means you have [music] to understand the dynamics in Chinese robotics right now. The prices, the speed, and what it [music] will do to the way you work. Because in 2 years, the situation will be totally different. The market will be much more mature. It will be very hard for you to [music] be one of the first mover. So, what do you think about Unitree? Is it worth $6 billion? Last year, it sold $237 million worth of robots. Do you think this is a fair price for a company like this, or are we early in a bubble? I am very interested in [music] what you think about this. Tell me in the comments. I read them, and I will talk about them in the next
[11:19] video. And if you find these stories useful, hit subscribe, give me a like. I do this every week here from China, so you do not miss it. And remember what I just said about Wang Xingxing working at DJI before he started Unitree. He is far from the only one. DJI is a crazy company and you have to pay attention to it. So, this story is called 200 million and nothing [music] to sell. It is about a company that raised 200 million US dollars with nothing to sell. The company is called Discover Robotics and beginning of July, a month ago, they raised over a hundred million dollars. The biggest single first round ever in Chinese consumer robotics. On August
[12:09] 3rd, less than a month later, they raised another hundred million. They have no product at all on the market and no sales. That is zero. Zero sales. So, why did anyone pay? Who are these investors? What are they thinking? Two things. The founders came out of Tsinghua University in 2023. Tsinghua is China's MIT or China's Harvard, whatever you want to call it. And the founders, Zhu Guoyu, he was a senior scientist and product manager at DJI. The CEO of this company ran the Mavic Mini at DJI and that is the little drone a lot of people own, right? And the head of AI ran DJI's intelligent flight system. So, what they
[12:57] want to build is a robot for your home. And I found an interesting quote of an investor and he said, "Why they paid the money? This is the only team in China that has already mass produced AI hardware by the million and can build the models and can also do the engineering." This team, Discover Robotics, they're not the only one. They're are the only company. So, I checked the numbers. Nearly 20 companies founded by ex-DJI people got funded in alone. That is a 5-year high. One of them raised $240 million in 6 months. Another one raised over 300 million. And
[13:47] almost all of them sit within 5 km of DJI headquarters in Shenzhen. They have a really nice headquarters. I go there every week, probably. And what they built is industrial robots, robot dogs for heavy loads, soft bionic robots, camera robots that follow you, flying systems. And one of them expects to deliver several thousand heavy-duty robot dog this year. And from the older ones, you may already own something. Bambu Lab, the 3D printers with a Kickstarter that was crazy, founded by the man who ran DJI's consumer drones. EcoFlow, also a very well-known brand, the power stations, founded by the man who built DJI's
[14:35] battery department. And Unitree, we just talked about. Wang Xingxing also worked for DJI before. Now, here's the part I love and I found very interesting. In Shenzhen, which is in southern China, one top investment firm rented a permanent office in the building across from DJI's towers. Not for a deal, just to watch who walks out. A department head says he's thinking about leaving, and he gets a $3 million valuation before he has even decided what to build. One manager from DJI had verbal commitments from three funds before he handed in his notice. Three different [music] funds. So, why DJI? And what does it mean for you and for your business? A drone basically is a robot
[15:24] that flies. We know that, right? Flight control, computer vision, embedded software, precision manufacturing. That is the same toolkit like a robot. And back to the engineers, DJI also trained a rare kind of engineer, [music] someone who understands the algorithm and the supply chain, and how to actually scale, how to build a million units >> [music] >> without the quality falling apart. Quality control, big thing at DJI. And the wins created a loop. So, Bambu Lab, EcoFlow, Unitree all worked. So, new now investors treat XDJI as a shortcut. The formula they say out loud in Shenzhen is DJI people plus Shenzhen supply chain plus AI plus a global market. And in
[16:13] Shenzhen, people talk about that formula like it's almost a guaranteed success. So, DJI has stopped being a company. Now, it actually turned [music] into a university. And what it means for you, you future robot supplier, probably did not exist 2 years ago. And the team behind it >> [music] >> has already shipped millions of units of consumer hardware. And that combination does not exist in Europe or in the US right now. So, next time you talk to a robotics company here, ask them one [music] thing. Did your founder or your CTO work at DJI? And if you don't know how to reach them, just ask me. I will ask them for you. [music]
[17:01] In the third story, let's talk about my favorite component. This story is about the hand of robots, and more precisely, the touch sensors in it. Half [music] a billion dollars for a sense of touch. The company that just got it is called Pacxini. They were founded in 2021, so they're one of the older ones, 5 years old already. And you notice something, like most of the companies in this show, they were founded in the last three to five years. And last week Pacxini raised another 140 million dollars. That brings them to about half a billion dollars in total. No company anywhere in the world has raised more money for touch sensors.
[17:50] Not in the US, not in Japan, and not in my home of Germany. Who are the investors? So, back in March, they were already valued at 1.4 billion dollars. And the new round was led by a semiconductor and consumer electronics giants, and together with the investment arm of Bank of China and a national fund. In the last 12 months, they used almost 1 million of their own tactile chips. So, this is already mass production. They are totally out of the prototype phase. Now, every humanoid you have seen can see. There's eyes and there's cameras, right? Cameras are cheap and vision models are really,
[18:38] really good now. But feeling is the part that nobody has solved yet. A robot that cannot feel has to be told exactly what it is picking up [music] and exactly how hard to squeeze. So, this is a problem that vision cannot fix. [music] The moment the hand closes around the object, the camera cannot see it anymore. The robot is blind at exactly the moment that is very important and crucial. So, Pacxini sensors read 15 different things at once. Six axis force, texture, how soft something is, [music] down to 100th of Newton. Their own humanoid, they're also building a humanoid. It's called
[19:27] Toro One. It carries 1,956 of these sensors. Now, the question, why is the money going there? Touch data makes the robot handle things it has never seen before. We We know that vision data and language data cannot do that on their own, right? So, they are building factories whose only job is to collect data. One is in Tianjin, which is around 1,000 km from here in northern China, and three more are going up right now. And the part I find really interesting, the founders came out of the Sukano lab at Waseda University in Tokyo, a very famous university, and was the lab that built
[20:17] the first humanoid robot in the world back in the [music] '70s. So, a Japanese research line turned into a Chinese company funded with Chinese money and now supplying the world. And here is what this means for you. Two humanoids can look identical on stage and be completely different machines. The difference is [music] mainly in the hands. This is something I run into a lot with my customers. So, my customers, they come to me and they say, "We have this or that problem, and we think that a robot from this or that brand is the solution." But, if you don't understand what these machines can actually do, you can spend a lot of money on a robot that
[21:05] does not work in your factory, in your hotel, or in your supermarket. So, you have to know which robot [music] is the right one for you. So, this is why I am here. If you want an assessment of which robot would work for you, just [music] let me know. I am a big fan of the smaller companies who have great products, great components, better vision, better tactile sensors, >> [music] >> um better hardware, better software, better brains than all the big names you know from the internet. So, if we have a look what happened this week in China, a listening that made engineers into millionaires, 200 million for a team with no product, half a billion for a sense of touch, 210 [music]
[21:53] million for a robot brain. All of this happened in China in just one week. And the big opportunity for all of us is getting into that supply chain because a supply chain [music] gets allocated once. Two weeks ago, I took a German automotive supplier to 12 different robotics companies in Suzhou, Hangzhou, >> [music] >> and Shenzhen, and Shanghai. And they all had the same problem. They need to make their components lighter, cheaper, and they need to get the heat down. So, if you have a solution for that, maybe you are an automotive supplier, they would love to talk to you. And the second one is the other direction, the obvious direction. You have to think about how you automate the processes in your own company and in your factory. Or even if
[22:43] you're a retailer, whatever, you need to think about automation and robotics. So, if either of these is your topic, the link to contact me is down in the description. Let's find a time and talk. And now, as promised, I will show you a video of a very cute robot. >> Sir? Sir? Sir, I told you not to water those plants. I told you to water the ones over there. Sir, sir, sir, sir. Thank you so much for watching this episode of Robotics News. Be sure to subscribe and leave me a comment. What do you think about the Chinese robotics companies? What do you think about Unitree's evaluation? Let me know in the comments and we see you next week.
Get These Insights Every Monday
Join 18,000+ professionals reading Asiabits. Free, every Monday, straight from Shanghai.
Subscribe Free →
