On 5h of November Agility Robotics opens on Nasdaq as AGLT with valuation $2.5 billion. Revenue for the first six months of 2026 is $174,000 (mostly from related parties). TechCrunch published both numbers in the same sentence and asked readers "are you buying?" - are you?
In the same week Xiaomi claims its humanoid worked a full 8-hour factory shift on its own. China's regulator kills SoftBank's $5.4 billion ABB deal on the exact same day that Agility lists. And last but not least - Physical Intelligence enters funding talks that could push its valuation to $20 billion, six weeks after launching a robot model API. Looking at all of this together, one question keeps coming back:
Who is actually building the infrastructure layer for humanoid robotics at scale and does anyone outside the top five companies know yet?
Before we start to wonder, here are the numbers worth remembering from this week:
- $2.5B - the worth of Agility Robotics pre-money valuation at Nasdaq debut ($465M raised via SPAC with Churchill Capital XI, Foxconn led the PIPE at $201M).
- $174K - Agility's H1 2026 revenue - mostly from related parties; nearly 100 Digit robots in commercial deployment, GXO expanding to 2 more US facilities in 2027.
- 8 hrs - the time that Xiaomi's humanoid worked autonomous factory shift at a sorting station in its EV plant - no independent verification published.
$2.5 billion on a category without a revenue model and nobody is pretending otherwise
First the mechanics - Agility went public via a SPAC merger with Churchill Capital Corp XI. Shareholders approved it on November 2nd, with more than 84% of them voting yes. After redemptions - 61.5% of Class A shareholders pulled out - total proceeds landed at roughly $465 million.
What Agility brought into the listing? Nearly 100 Digit robots in real commercial deployments. GXO moved over 100,000 totes using Digit at one facility and committed to two more US sites in 2027. The commercial terms of those contracts remain undisclosed. The $174,000 is what made it onto the balance sheet.
So what exactly is the market paying $2.5 billion for? Clearly not the revenue. My read: it's a bet that three things happen simultaneously.
First one, humanoid robots become a standard CAPEX line item in warehouses and factories within around five to seven years.
Second, Digit is the right platform bet in the US market, partially because access to Amazon's real warehouse environments gave Agility training data nobody else had.
Third, being public gives Agility capital fast enough to stay ahead of Chinese manufacturers closing in on the same use cases at half the price. All three have to land. At the same time.
Unitree listed in Shanghai in August. Agility listed in New York in November. The world's two largest markets each now have a pure-play humanoid stock. The public market isn't validating the technology. It's pricing the race.
Xiaomi's humanoid 8-hour shift: what the company said and what it didn't
Xiaomi published footage of its humanoid at a side-cover sorting station in its EV factory - on-screen timer running, autonomous anomaly correction described. Eight continuous hours, the company says. Put it in context: March was a 3-hour trial at 90% task completion. July, after a four-month "internship," was 98% success rate at a specific nut-tightening workstation. The progression is real.
What was not published is whether the run included charging stops or battery swaps, how often operators intervened, what the error rate actually was and what throughput looked like compared to a human on the same station. There is no independent verification. The claim stands on Xiaomi's word.
And yet - the trajectory is hard to ignore. March: 3h, 90%. July: 98% after four months. And on November: claimed full shift. Even if the numbers are rounded optimistically, Xiaomi is compressing its deployment timeline faster than Western companies with longer track records. Lei Jun has committed to large-scale humanoid deployment across Xiaomi's production facilities within five years. Looking at what they've published over the past eight months, that timeline is not an obvious bluff.
Briefly: China blocked the ABB deal on the day Agility listed
On November 5, China's National Development and Reform Commission rejected SoftBank's $5.4 billion acquisition of ABB's robotics division. The timing is probably coincidence. The significance is not. ABB is one of four dominant industrial robot manufacturers globally - alongside FANUC, KUKA, and Yaskawa - with deep deployment inside Chinese factories at scale. A SoftBank-owned ABB would have put a Japanese-controlled entity inside Chinese manufacturing infrastructure at exactly the moment China is pushing hardest to build domestic humanoid capacity. The block keeps that asset independent. SoftBank has not said what comes next - a renegotiated structure, a different buyer, or something else entirely.
During one week, there were three seperate events. The real question isn't whether Agility deserves $2.5 billion at $174,000 in revenue. The real question is what it means that this is the moment public markets decided to price this category at all. Five years from now, someone will look back at November 2026 and either point to it as when the race became official or as when the hype peaked. We don't know which yet. But the scoreboard is up.
Briefing prepared by Klaudia from xBerry - a tech company based in Poland at the intersection of Physical AI and industrial operations.
Top comments (0)