Humanoid raises £114M and becomes Europe's first humanoid-robot unicorn

Humanoid, a London robotics company, has raised $152M, or roughly £114M, in a Series A round that values it at $1.35B, or around £1bn. The company said the raise makes it Europe’s first pure-play humanoid-robotics unicorn. The round was led by Prime Movers Lab, the venture firm behind the US humanoid company Figure AI, with Bosch, Schaeffler, Fubon Financial Holding Venture Capital, and Bernard Arnault’s family office Aglaé Ventures also taking part.

For a Series A, the size is the story. Most companies raise this kind of money years into their life, if at all. Humanoid says the round takes its total funding to about $270M and will pay for commercial deployments of its wheeled humanoid robots, more manufacturing capacity, and further work on its in-house AI system, which it calls KinetIQ.

The bet the money is backing is a specific one. Rather than build a robot that walks on two legs, the hardest and least reliable part of humanoid robotics, Humanoid puts its upper-body robot on a wheeled base aimed at warehouses, factories, and other flat industrial floors. It is a pragmatic design choice: give up the ability to climb stairs in exchange for a machine that is cheaper, steadier, and closer to being useful today. The tie-up with Bosch matters here, because a robotics startup’s hardest problem after the technology is making the thing at scale, and Bosch is set to manufacture the robots.

For the UK, it caps a striking month for home-grown hard tech. It follows CuspAI’s £335M raise for AI materials discovery, and together the two rounds point to a shift in where UK AI money is going: not only into software and models, but into companies that marry AI with physical hardware, which is capital-intensive and slow but far harder for rivals to copy. A London company reaching a $1.35B valuation on a Series A also gives the UK a rare late-stage robotics name to point to, in a field the US and China have dominated.

What makes these robots “AI-powered” rather than just automated is worth unpacking. A traditional industrial robot repeats a fixed, pre-programmed motion in a cage. A humanoid robot has to perceive a messy, changing environment and decide what to do in it, which leans on computer vision, the field of AI that interprets images and video, and on machine learning, where a system learns behaviour from examples rather than from rules written by hand. Humanoid’s KinetIQ platform is the layer that turns what the robot sees into how it moves and grips. How well that generalises from a demo floor to a real, cluttered warehouse is the open question behind every company in this space, and the one the next few years will answer.