Key points
- Agility Robotics offers the most direct publicly accessible US humanoid wager through its planned CCXI SPAC merger. Digit is already performing paid jobs for GXO, Toyota, and Schaeffler, and Agility reports more than $300 million in contracted orders.
- The catalyst is China. Unitree's Shanghai IPO drew about $1.2 trillion in orders, oversubscribed up to 8,288x, and with the US behind, expect Washington and Silicon Valley to spend heavily to catch up.
- The entry price is not cheap. CCXI trades near $16, implying an Agility valuation between $4 billion and $5 billion before the transaction closes, so any position should remain modest.
- Lower-risk ways in are the names already making money, Teradyne (TER) and Nvidia (NVDA), plus Tesla (TSLA) for Optimus or a basket ETF like BOTZ.
A Chinese manufacturer of robot dogs and humanoids has completed one of the most remarkable IPOs in years. Unitree offered its stock on Shanghai's STAR Market on August 6 at 150.8 yuan, or about $22 per share, giving the company a valuation near 61 billion yuan, approximately $9 billion. Demand then surged. Retail buyers entered roughly 9.8 million subscriptions valued at about 8.1 trillion yuan, close to $1.2 trillion, for a transaction that raised less than $1 billion. Orders exceeded the shares available by up to 8,288 times, and an online applicant had only about a 0.018% chance of receiving an allocation.
Read that again. Roughly $1.2 trillion in demand for a company that did about 1.7 billion yuan, or $240 million, in revenue last year and shipped a little over 5,000 humanoid units. Normal valuation discipline does not explain that level of demand; speculative mania does. The important takeaway is what sits beneath the frenzy: China is committing heavily to robotics while the United States is still trying to catch up.
China has committed, and America is trying to close the distance
No pure humanoid-robot company had previously listed on a mainland Chinese exchange before Unitree. Beijing has also made "embodied AI" a national objective, much as it did with electric vehicles ten years ago. China brings together factories, an established supply chain and a government prepared to finance domestic robot production. Unitree already sells a walking humanoid for a small fraction of the cost of an American-built alternative. A country falls behind when its competitor can build the hardware more quickly and cheaply with government subsidies supporting the effort. It is better to acknowledge that reality directly.
Here's the part we'd bet on, though. The US is not going to sit back and watch China own the next platform after the smartphone. Washington already restricts the most advanced Nvidia chips from shipping to China, and robotics sits right at the intersection of AI, manufacturing, and defense, which is the exact combination that gets both political parties spending. The Pentagon wants autonomous systems. Silicon Valley is pouring private money into humanoid startups. When a country decides it can't afford to lose a technology, the money shows up, and it shows up in the tens of billions. That's the tailwind behind this whole trade.
The most direct US play: Agility Robotics (CCXI)
The practical problem is finding a way to invest in the American side. Figure and many other leading US humanoid developers remain private, leaving public investors without access. Agility Robotics is the main exception because it plans to merge with Churchill Capital Corp XI (CCXI).
Agility is the highest-upside name on this list, and it's not particularly close. Its Digit humanoid is already doing paid work in real warehouses, which almost none of its rivals can say. Digit has moved more than 100,000 totes at GXO's facility in Georgia since 2024, the most documented commercial humanoid hours of any robot on earth, and it's now logged over 65,000 operating hours across nine committed facilities. The customer list reads like a who's who: GXO signed the industry's first commercial humanoid rental contract, Toyota put seven Digits to work at the plant that builds the RAV4 in Ontario, and Schaeffler and Mercado Libre are both on board. Agility says it's already sitting on more than $300 million of multi-year Digit v5 orders, and it builds the robots at RobotFab, its Oregon factory designed to eventually turn out up to 10,000 units a year. This is a real order book from real customers, not a rendering.
Now the catch, because there's always one. You're paying up for it. The deal values Agility at about $2.5 billion at the $10 reference price, but CCXI trades closer to $16, and against roughly 325 million pro-forma shares that puts the real price tag somewhere between $4 billion and $5 billion, before the merger has even closed. As of mid-August the deal hadn't filed its full S-4 and there's no shareholder vote scheduled yet, so this is still a pre-close SPAC with all the risk that carries. The encouraging tell is who's funding it: the PIPE is led by Foxconn, the company that assembles the iPhone, which knows a thing or two about manufacturing at scale. The upside here is enormous if humanoids work, the market cap is still small next to that dream, and the way to own it is a small position you can stomach watching bounce around. Buy it like a venture bet, not a savings account.
Less volatile exposure to the same trend
Anyone who finds a pre-close SPAC too aggressive can instead hold companies that profit from robotics regardless of which humanoid platform prevails. Teradyne (TER) is the most straightforward example. Although known mainly for testing semiconductors, Teradyne also controls Universal Robots and Mobile Industrial Robots. That segment recently recorded its first $100 million quarter, an increase of 33% from the year before. Chief executive Greg Smith said the unit had set a revenue record for the second consecutive quarter, with AI once again driving the result, and noted that more than 60% of its sales are now connected to AI. Those figures represent equipment being delivered now rather than forecasts in a presentation.
Nvidia (NVDA) supplies the underlying tools for nearly the entire industry. The company does not manufacture humanoids. Instead, it provides the processors that operate them, simulation platforms used for training and foundation models that allow a robot to learn how to grasp a box without engineers coding each motion individually. For more than a year, Jensen Huang has characterized physical AI as a multitrillion-dollar opportunity. Nvidia benefits because nearly every credible robotics developer in China or the United States can become a customer. The AI compute demand powering the data-center boom is what makes robots possible in the first place.
Then there's Tesla (TSLA), the mega-cap version of the bet. Its Optimus humanoid is running simple tasks inside Tesla's factories, and Musk has gone as far as calling Optimus a potential $10 trillion business. We'd take that with a boulder of salt. This is the same guy who promised roughly 10,000 robots built in 2025 and missed it, and each Optimus still costs more than $60,000 to make. Tesla is the highest-profile way to own this theme and the one most likely to disappoint on timing.
It's still extremely early, and how we would approach it.
If picking one winner feels like a coin flip, a basket takes the guesswork out. The Global X Robotics and AI ETF (BOTZ) and the ROBO Global fund (ROBO) spread across dozens of automation names, and ARK's ARKQ leans hardest into Tesla and autonomy. One honest warning: these are global funds, not pure US bets. BOTZ's biggest holdings include Japan's Keyence and Fanuc and Switzerland's ABB alongside Nvidia, so you're buying the whole robotics supply chain, not a red-white-and-blue humanoid index. A clean US-only humanoid ETF doesn't exist yet.
Our take: this is a real, decade-long shift, China has the early lead in robotics, and the money in the US will chase it hard to close the gap. Start small, expect the charts to look like a heart monitor, and remember that being early and being wrong can look identical for a long time before they don't. If you're building a position from scratch, the order you put money to work in matters more than which robot stock you buy first.
This is general market commentary and opinion, not investment advice. I am not a financial advisor. Prices are intraday on August 14, 2026 and will change. Early-stage robotics and pre-merger SPAC bets are high risk, and you can lose money. Always do your own research and consider speaking with a licensed financial professional before making any investment decision.
