Tesla Optimus vs Unitree: China Already SOLD Thousands Of Humanoid Robots
Publicado no IA Dots ·
AI Signaldot-artificial_intelligencedot-figuredot-humanoid_robotsdot-teslaidioma-enInglêsRobótica e humanoidesyt-ZiOgl3i27kM
Vídeo em Inglês, publicado pelo canal Four Sides.
About this content: The summary, key points and topics below are generated automatically from the video transcript. They reflect the content, opinions, claims and predictions presented in the video and do not represent a position, validation or fact-check by IA Dots.
Video summary
The video analyzes the global humanoid robot race, comparing production volume and sales led by Chinese companies like Unitree and AgiBot, practical deployments by American startups such as Figure AI and Agility Robotics, Tesla's current development status with Optimus, and the impact of new U.S. federal regulations on foreign-made robots.
Key points
- Chinese companies like Unitree and AgiBot dominate global humanoid robot shipments, selling thousands of units primarily to labs and universities.
- American startups like Figure AI and Agility Robotics lead in real-world deployment, with robots performing paid tasks in automotive plants and warehouses.
- Tesla's Optimus remains in the research and development phase inside Tesla factories, with production numbers unconfirmed despite ambitious public targets from Elon Musk.
- The Federal Communications Commission added foreign-made mobile robots to its covered list, requiring special approval for new models built outside the U.S.
- Chinese humanoid manufacturers face major barriers entering the U.S. market due to new regulations, while Western startups and investors gain protected time and market advantage.
Topics
Humanoid RobotsChina RoboticsTesla OptimusFigure AIUnitreeRobot ManufacturingFCC Regulations
Video transcript
Transcript from available captions; it may contain errors.
According to the research firm Omdia, about 13,000 humanoid robots shipped worldwide last year. Three Chinese companies alone shipped more than 10,000 of them. And while those robots were shipping, the man most of the West treats as the face of humanoid robots said this about his own. It's not in usage in our factories in a material way. It raises a question. If China is shipping thousands and BMW already has a robot on its line, why does almost every story about the robot race still start with Tesla? You've probably heard the version that goes, Tesla already builds the cars and the AI. So Tesla will build the robot and everyone else is playing catch up. Elon Musk has even put a number on it. In September 2025, he posted that about 80 % of Tesla's value will be Optimus. For a lot of people, that post is the whole picture. But the reality is this race has been running for years and the scoreboard looks nothing like the pitch. So today, four sides. Who's shipping robots? Who's putting them to work? Where Tesla really is and who gets to decide who wins? That last one isn't an engineer. Start with the price tag. This is the Unitree G1, made in Hangzhou in eastern China. When it launched in 2024, it cost $16,000 and the prices dropped since. By August this year, Unitree's own store listed the base G1 at $13,500 with shipping and import duties on top. That's still less than most new cars. In July 2025, at the World AI Conference in Shanghai, Unitree went lower. It showed the R1 starting at $5,900. The cheap ones aren't factory workers though. One reviewer noted the R1 doesn't even have working hands and most of these robots end up in labs and universities. Hold on to that because selling to labs is how Unitree got to its next number. When Unitree filed to go public this year, it had to open its books. In 2023, it sold five humanoid robots. In 2025, it sold 5,215. About 70 % went to universities and research institutions. So a lot of the next wave of robotics engineers is learning on a Chinese robot. Unitree makes money doing it too. Its prospectus showed a profit of around $41 million last year. In August, it listed on Shanghai's star market and raised about $900 million. On the first day, the stock closed 460 % higher. And Unitree isn't even number one. Omdia ranked AgiBot, another Chinese company, first in 2025 with about 5,100 units. That count includes some wheeled robots. So the two are roughly tied. UBTech from Shenzhen came third with about 1,000. That's who's shipping. Whether those robots have real jobs is another question. And on that one, the leader isn't Chinese. Spartanburg, South Carolina is home to BMW's biggest plant in the world. This is Figure O2, from a California startup called Figure AI. It worked in the body shop, picking up sheet metal parts and placing them for welding, 10 hours a shift, 5 days a week. By BMW's own count, it helped build more than 30,000 BMW X3s over 10 months. It moved more than 90,000 parts and logged about 1,250 operating hours. Figure's founder, Brett Adcock, put it this way. Our 11 -month deployment of Figure O2 proved that humanoids are no longer lab experiments. BMW says 10 months, and Adcock says 11. Neither has said how many robots were on the line, and the reporting points to one, maybe two. So this was a small pilot. BMW kept it going anyway. In June this year, Figure's next model, Figure O3, moved into Spartanburg's logistics hall. Now, it sorts loose parts from bins into trolleys that roll to the assembly line. BMW isn't betting on one supplier, either. In Leipzig, Germany, it brought in a humanoid called Aeon from Hexagon to help with battery production. Figure, by the way, was valued at $39 billion last September. There's more of this in American warehouses. Agility Robotics says its robot, Digit, has moved more than 100,000 totes at a GXO warehouse in Georgia on a paid contract. Two sides in, China leads in robots sold, and American startups lead in robots doing paid work, which leaves the company everyone talks about. To see where Tesla is, start with what it promised. At Tesla's shareholder meeting in June 2024, [Elon Musk] And my prediction is next year we'll have over a thousand, [Elon Musk] maybe a few thousand, Optimus robots working at Tesla. On the earnings call in January 2025, the internal plan was roughly 10,000 robots built that year. Then came the same call one year later, in January 2026. We are still very much at the early stages of Optimus. It's still in the R &D phase. When he was asked how many Optimus robots Tesla actually has, he didn't give a number. Now, Tesla is building something. This year, it shut down the Fremont lines that made the Model S and Model X, the cars that made Tesla famous, to make room for Optimus. On the April earnings call, Musk said Optimus has about 10,000 unique parts on a production line that's entirely new. He called the production rate this year literally impossible to predict. Tesla's long -term target is a million robots a year. A target isn't a shipment though, and the first robots off that line aren't going to customers. Tesla says they'll stay inside Tesla, doing simple tasks in its own factories. So, Tesla has no price, no order page, and no outside customer for Optimus. And by Musk's own words, most of its stock price rides on it. Then, in July, Washington stepped into the race. On July 28th, the Federal Communications Commission added foreign -made mobile robots, including humanoids and robot dogs, to what it calls the covered list. In plain terms, a new robot model built outside the U.S. can't get the FCC approval it needs to be sold here, unless the Department of War signs off on it. Robots that were already approved can still be sold. The notice names no country. It covers robots made anywhere outside the U.S., but most of the world's humanoids are made in China, so Chinese makers take the biggest hit. The cheapest, best -selling humanoid makers in the world can't bring their next model to one of the biggest markets in the world without Washington's permission. Optimus is no better than it was in June. It just has fewer rivals here. So who benefits? American robot makers get time, and they need it because the price gap is huge. None of them has publicly matched a $5,900 robot. Their investors benefit, too. Figures valuation, agility plans to go public, and a large share of Tesla's stock price all lean on the idea that the West can win this. And who pays? American labs and universities, for one. The buyers who made Unitree a business now have fewer options and pricier ones. China hasn't won either. Most of its robots still do research and demos, and UBTech, the number three, lost around $104 million last year. Nobody is close to the finish line yet. When people say Tesla is winning the robot race, they usually mean Tesla is winning the robot's story. That story is worth a lot. It helps move a stock worth more than most car makers combined. The robots that actually sell are mostly Chinese and cheap, and they get cheaper every year. The ones doing paid work mostly come from American startups, in small numbers, in places like Spartanburg and a warehouse in Georgia. And the most famous robot in the world is still, in its maker's words, in the R &D phase. Tesla could still win this. It has money, factories, and a habit of showing up late and then showing up big. But right now, Tesla's talk about robots is far ahead of its robots. Last year, Unitree sold 5,215 humanoid robots. Tesla hasn't said how many it built.
Assistir ao vídeo original no YouTube
Assuntos: Robótica e humanoides, Robótica e humanoides. Publicado no YouTube em 2026-10-18.
Seleção automática por assunto e idioma. O vídeo pertence ao canal de origem.