Let's set the scene: humanoid robotics is currently experiencing the kind of funding frenzy that makes even seasoned VCs blush. A Shenzhen-based startup just pulled in $735 million at a $3 billion valuation. Apptronik—backed by Google, Mercedes-Benz, and John Deere—closed a $935 million round that pegged its value north of $5.5 billion. And Figure AI, with the audacity only a San Jose startup can muster, self-reported a $39 billion valuation on a $1 billion Series C. In an industry where "real-world deployment" often means a carefully choreographed demo for a YouTube audience, these numbers are eyebrow-raising.

Against that backdrop, Agility Robotics CEO Peggy Johnson sounds almost suspiciously sane.

Agility just announced it's going public through a merger with Churchill Capital Corp XI, a SPAC operated by financier Michael Klein. The deal pegs Agility's value at roughly $2.5 billion—a figure that feels almost modest by current humanoid standards—and is expected to generate over $620 million in gross proceeds, which the company claims is the largest capital raise in humanoid robotics history. The merger still needs SEC review and shareholder approval, so don't start refreshing your brokerage app just yet. But if it closes as expected, Agility will become something genuinely new: the first pure-play humanoid robotics company trading on public markets.

That's actually significant. Not because the stock is guaranteed to do anything useful, but because public markets demand financial transparency that private rounds simply don't. For the first time, retail investors—and the rest of us curious about what these businesses actually look like under the hood—will get real numbers instead of press release vapor.

A CEO Who Doesn't Promise You a Robot Butler

Johnson's résumé is worth understanding, because it shapes how she runs Agility. She helped engineer Microsoft's $26 billion acquisition of LinkedIn, then took the helm at Magic Leap—the augmented reality headset company that burned through billions of investor dollars before dramatically pivoting to enterprise applications. She's seen what happens when hardware hype outruns hardware reality. It shows in how she talks.

Asked why a SPAC instead of a traditional IPO roadshow, Johnson frames it as a first-mover play. Agility is "an acceleration story and a timing story"—be the first humanoid company on public markets, capture investor appetite before competitors do the same. It's a defensible strategic rationale, though it conveniently sidesteps the fact that SPACs also skip the pricing scrutiny that comes with a traditional IPO process. As for the SPAC horror show of 2021—when dozens of companies went public through blank-check vehicles and subsequently collapsed—Johnson's response is essentially: we'll be different because we'll actually execute. Bold claim, but at least she's not pretending the track record doesn't exist.

The Robot Itself: Deliberately Unfussy

Agility's flagship product, Digit, is a bipedal humanoid designed around a specific, unsexy thesis: move heavy stuff reliably in spaces built for humans. It stands about 5'9", weighs around 160 pounds, and features a set of reverse-bend "bird legs" that let it reach from floor to overhead shelving without knocking into warehouse racking. The hands have two fingers and two thumbs, optimized for gripping heavy plastic totes—not for playing piano. Agility's founders, Johnson explained, had zero interest in biomimicry for its own sake. They wanted something that works in a real warehouse, not something that looks good in a press release.

This is the right instinct, and it's worth saying clearly: most humanoid robot demos are carefully staged to hide the catastrophic failure modes. Agility has actually put Digit in real distribution centers for real customers—GXO Logistics, Amazon, Toyota Motor Manufacturing Canada, Schaeffler, and Mercado Libre are named in the pipeline. Johnson says there are over $300 million in booked, multi-year revenues representing roughly 1,000 robots on a robots-as-a-service model, where customers pay monthly rather than buying the hardware outright. That's a subscription business layered on top of a hardware business, which is how you build recurring revenue—and also how you obscure unit economics if you're not careful. Public filings will tell us more.

The LLM Layer: Smart About What It Can't Do

Agility describes itself as "LLM-agnostic," pulling from models including Anthropic's Claude and Google's Gemini to handle what Johnson calls the semantic layer—converting natural-language instructions into robot behavior. The demo she cited: engineers scattered various types of trash on the floor, told Digit to "clean up this mess," and the robot correctly sorted and binned everything, including identifying bubble wrap as non-recyclable. It's a good demo. Whether it generalizes to novel environments without human oversight is the harder question.

Johnson is candid about where the real proprietary value lies, and it isn't in the LLMs. "The LLMs had the entire internet to train on," she said. "When you think about the physical AI of humanoids—that doesn't quite exist yet." The physical layer—balance, locomotion, manipulation in unpredictable environments—is where Agility claims its genuine edge: over a decade of real-world deployment data that competitors simply don't have. If that data lake claim holds up under scrutiny, it's the most defensible moat in the business. Physical AI training data is genuinely scarce, and unlike text scraped from the internet, nobody's giving it away for free.

What the IPO Actually Tells Us

Founded in 2015 as a spinoff from Oregon State University and headquartered in Salem, Oregon, Agility has been grinding away at this problem longer than most of its flashier competitors have existed. The $2.5 billion valuation reflects that measured positioning—this isn't the "we'll disrupt everything" pitch deck energy of a $39 billion startup. The public listing will force disclosure of real financials: cost per robot, gross margins, burn rate, the actual economics of that RaaS model. That transparency is either going to validate the thesis or expose the gap between the story and the spreadsheet.

Either outcome is useful. If Agility's numbers hold up, it sets a credible benchmark for the entire sector—and provides cover for the companies currently raising at valuations that would make a WeWork executive blush. If the numbers disappoint, the sector gets a necessary correction before even more capital gets incinerated on robots that mostly live in controlled demos.

Johnson isn't promising a robot in your home. She isn't promising to replace all warehouse workers by 2027. What she's promising is robots that can move totes in a distribution center—reliably, at scale, for paying customers. In humanoid robotics right now, that's practically a radical act of modesty.