26 million humanoid robots a year: Berg Insight maps out the growth curve
Key Takeaways
- Berg Insight forecasts humanoid robot shipments will grow from 16,000 units in 2026 to 26 million annually by 2040: a 63.7% CAGR
- Market value is projected to grow from $890 million to $554 billion over the same period
- More than 100 companies are now developing full-size humanoid robots
- Competition is shifting from hardware performance toward software, embodied AI, and scalable manufacturing
- NVIDIA, Microsoft, Google, Amazon, Hyundai, and OpenAI are all now strategic investors in the space
There's an important number in Berg Insight's new humanoid robot forecast that is a little small: just 16,000. That's where the industry stands in 2026 ... 16,000 units shipped annually, a market worth just $890 million, and there's a long way to go.
To get from 16,000 to 26 million by 2040 requires a 63.7% compound annual growth rate, sustained for 14 straight years. To turn $890 million into $554 billion requires roughly the same. According to IoT research firm Berg Insight's newly published report, that's exactly what's coming.
Probably. Frankly, this is conservative by some estimates.
But forecasts this far out are always educated bets, not certainties.
What 63.7% CAGR actually looks like
A lot of technology forecasts cite big CAGRs on small bases: impressive numbers that dissolve under scrutiny. 63.7% sounds aggressive until you remember what smartphone shipments looked like from 2007 to 2014, or what EV adoption curves looked like from 2012 to 2022.
Fast compounding on small bases is exactly how new platform markets work.
The industrial math is also straightforward. Global manufacturing employs roughly 300 million people; global logistics another 300 million. Elder care offers work for hundreds of millions more. A humanoid robot that reliably performs physical labor at even a fraction of human cost unlocks enormous economic value. The constraint isn't demand; it's whether the robots can actually do the work consistently and cheaply enough to justify purchase.
"The humanoid robotics industry has reached an important commercial inflection point," says Julia Wahlen, market analyst at Berg Insight. "For the first time, we are seeing several companies simultaneously scaling production, establishing dedicated manufacturing facilities and deploying humanoid robots in real-world commercial environments."
That simultaneity is new.
A year ago, only a handful of companies had real commercial deployments. Now Agility Robotics is in Amazon warehouses, Boston Dynamics is heading to Hyundai plants, Figure is in BMW facilities, and Tesla's Optimus is starting assembly line work.
The demo era is ending. The work era is beginning.
Where competition is actually happening
Here's what's changed in the last 18 months: the hardware problem is largely solved at the high end, and it's being commoditized at the low end by Chinese manufacturers. Unitree's G1 is $13,000.
Chinese university spinouts are now shipping humanoids at $1,350.
The actuators and motors are known technology. They're going to get better, yes, but more importantly, they're going to get cheaper.
"Competition is increasingly shifting from hardware performance alone towards software, embodied AI, and scalable manufacturing," notes Emilia Rickard, also of Berg Insight. That framing aligns with what the leading players are actually investing in — foundation models, dexterous manipulation, generalist task learning, and the ability to deploy across different environments without months of reprogramming.
The 100+ companies currently developing full-size humanoids aren't all going to survive to 2040. Markets this size tend to consolidate. But the competitive pressure they're generating is accelerating the pace of improvement across the board.
Strategic capital and what it signals
The Berg report specifically calls out the wave of strategic investment that's now embedding humanoid robotics into the broader tech stack: NVIDIA, Microsoft, Google, Amazon, Hyundai, and OpenAI all have significant positions in the sector.
These aren't financial bets. They're infrastructure bets.
NVIDIA is building the simulation and training platform. OpenAI is building the foundation models. Amazon and Hyundai are building the deployment pipelines. When the biggest technology companies in the world are treating humanoid robots as core infrastructure rather than interesting adjacent markets, it changes the probability distribution on outcomes.
The cellular connectivity angle in Berg's report is also worth noting: subscriptions tied to humanoid robots are forecast to grow at 84.9% CAGR, from 5,000 in 2025 to 50.6 million by 2040. That's a connectivity backbone that doesn't exist yet, and a significant opportunity for network operators and edge computing providers to get ahead of.
The risks that could break this forecast
Regulation is the most obvious wildcard.
Humanoid robots in workplaces raise questions about liability, safety certification, labor displacement, and data privacy that governments are only beginning to grapple with. A major accident at a high-profile deployment — the robotics equivalent of a self-driving fatality — could trigger regulatory chilling that delays adoption by years.
The demand-reality gap is the subtler risk.
China is producing humanoids at scale right now, but producing for whom? The honest answer is that real commercial applications with proven ROI are still relatively narrow: parts sequencing in automotive, some warehouse operations, a few agricultural pilots.
Building the supply chain for 26 million units requires demand that isn't fully proven yet.
Still, the direction of travel is clear. From 16,000 to 26 million is a 1,625x increase in 14 years. The models that got there in smartphones, EVs, and solar panels all looked similarly implausible from the baseline. The question isn't really whether the humanoid market gets large. It's whether the companies placing bets today are the ones still standing when it does.
Sources
Written by
I tell stories with data and build startups. I've raised angel and VC funds, led software companies and startups, and chronicled the rise of the mobile economy. Currently I write for Forbes, host several podcasts, and am starting up HumanoidMegaHub. I was TUNE's mobile economist and Singular's VP of Insights, built the research team at VentureBeat, and managed teams creating software for partners like Intel and Disney. In past lives I built websites, e-learning platforms, and mobile apps, and, a long time ago, was named to Folio's top 100 of the media industry's "most innovative entrepreneurs and market shaker-uppers.”