Musk's 10 Million Robots: Who Is Actually Getting Paid?

Tesla shut down its Model S and Model X line, tying the fate of its Fremont factory to a product that has yet to prove itself. The old line produced about 40,000 luxury EVs a year; the retooled one aims for 1 million Optimus humanoid robots annually.

That is a 25-fold capacity increase—but Musk has not named the buyers.

That is the crux. Everyone debates whether humanoids will change the world, but the sharper question is: when Musk says he will build 10 million robots, who supplies the parts and who gets paid?

A contrarian call: the surest money in this wave is not Tesla itself, but the Chinese suppliers forced to expand alongside Musk.

The reason is simple. Even Tesla cannot make its own cameras, sensors, chips, motors, reducers, ball screws, and magnetic materials. J.P. Morgan estimates a humanoid's operating cost at $10–12 per hour; two robots match one human worker, whose labor costs about $30 per hour. To push costs that low, the supply chain must be extremely mature.

And the world's most mature robot-component supply chain is in China. According to Caijing, by July 2026 several Chinese suppliers had received orders for hundreds of third-generation Optimus components. Tesla's capacity guidance: 150 units a week in July, 300 in August, 1,000 in September, with some suppliers asked for 1,500 a week. One supplier said 70% of its factory capacity is now dedicated to Optimus. These are not intentions—they are firm orders.

Sanhua Intelligent Controls supplies thermal management for Optimus; Tuopu Group started with linear actuators and expanded into rotary actuators and dexterous-hand motors, completing multiple sample rounds. Leader Harmonious Drive's harmonic reducers and Hechuan Technology's servo systems are also on Tesla's list. CITIC Securities reads the Fremont line's shift to robots as a substantive pre-mass-production move.

A second call: this supply-chain rally will be more concentrated than the EV wave, with fewer winners. In the EV era, battery, motor, and electronic-control suppliers were relatively fragmented—CATL, BYD, LG, and Panasonic all got a slice. But humanoid core components—actuators, reducers, ball screws—have higher technical barriers and longer qualification cycles; once inside the T-chain, switching costs are extremely high. Sanhua and Tuopu hold core actuator positions. Mainstream market view names Xinjian Transmission, Sanhua Intelligent Controls, and Tuopu Group as Tesla's three core Optimus suppliers.

That means latecomers will struggle to get in. Tesla procures components under a framework of 100,000 intended orders and asks suppliers to expand to that capacity. Expand, and the orders are yours; don't, and someone else does—you are out.

A third call, on timing: real earnings will arrive in 2027, but the stock-pricing window is now. T-chain listed companies still deliver mostly small batches; large orders take time. Musk himself admits Optimus is Tesla's hardest product to mass-produce, with capacity following an S-curve—flat and prolonged at first.

Short term, don't expect pretty financials. But capital markets never wait for financials. Changjiang Securities says the robot industry has reached a stage of separating substance from hype, with strategy, channels, and technology deciding who enters the T-chain. Suppliers that have already proven themselves in Tesla's auto business are copying that relationship into robotics.

At the G20, Musk said the world will have 1 billion humanoid robots within a decade, each five times more productive than a human. Whether the number is right matters less than that he believes it—and is using hard orders to make suppliers believe it too. No one knows if the 10-million goal will be met. But one thing is certain: whether final output is 100,000 or 10 million, suppliers that secured orders early, expanded capacity, and locked key positions are already getting paid.