Figure AI: The Startup That Raised $675 Million to Build Your Robot Coworker

$675 million before selling a single robot

In February 2024, Figure AI announced it had raised $675 million in a funding round that included Microsoft, OpenAI, Intel, Jeff Bezos, and Nvidia. The company was founded in 2022. At the time of the raise, it had not shipped a single commercial robot.

That funding round tells you something important about the moment we're in. The most sophisticated investors in the world — people with deep technical knowledge and access to every deal — looked at a two-year-old humanoid robotics startup and decided it was worth betting hundreds of millions on. That's not hype. That's signal.

So what is Figure AI actually building, and why does it have some of the world's most serious investors so convinced?

The founding vision

Figure was founded by Brett Adcock, who previously co-founded Vettery (sold to Adecco) and Archer Aviation (electric air taxis, now publicly traded). Adcock is a serial entrepreneur with a track record of building in hard, capital-intensive industries — not a researcher who stumbled into a company, but someone who deliberately chose humanoid robotics as his next target.

The founding thesis is straightforward: there are tens of millions of jobs in warehouses, factories, and logistics that are physically demanding, repetitive, and increasingly hard to staff. Humanoid robots — machines that can work alongside humans in spaces designed for humans, using tools designed for humans — are the natural solution. And the technology has finally reached the point where building them is possible.

Figure's stated goal is to deploy humanoid robots into the workforce to address labour shortages and handle dangerous or unpleasant work. It's an unsexy pitch dressed in exciting hardware.

Figure 01: the first robot

Figure's first robot, called Figure 01, is a full-size humanoid — roughly human height and weight — designed specifically for industrial work. It has two arms, two legs, and hands capable of fine manipulation. It's built to operate in existing industrial environments without requiring those environments to be redesigned around the robot.

This last point matters more than it might seem. Many industrial robots are fixed in place, requiring purpose-built facilities around them. A humanoid robot that can walk into an existing warehouse and start working alongside humans is a fundamentally different product — one that removes the enormous capital cost of facility redesign.

In early 2024, Figure released a video showing Figure 01 having a fluid spoken conversation with a human, identifying objects on a table, and responding to requests — picking up items, explaining its reasoning, and demonstrating a level of natural interaction that felt qualitatively different from earlier robot demos. The video was produced in collaboration with OpenAI, whose language models were powering the robot's conversational and reasoning capabilities.

The BMW deal

In January 2024, Figure announced a commercial agreement with BMW to deploy Figure 01 robots in BMW's Spartanburg, South Carolina manufacturing plant — one of the largest BMW facilities in the world. This was the first commercial deployment agreement for a Figure robot.

The Spartanburg plant produces around 1,500 vehicles per day and employs thousands of workers. The tasks Figure robots are being deployed for involve moving parts, loading components, and handling materials — physically demanding work that's difficult to automate with traditional fixed robots due to the variability involved.

The BMW deal matters for two reasons. First, it's real — an actual commercial contract with one of the world's largest manufacturers, not a pilot program or a letter of intent. Second, BMW is a sophisticated buyer with extremely high standards for manufacturing quality and reliability. Their willingness to deploy Figure robots is a meaningful endorsement.

Figure 02: the second generation

In August 2024, Figure unveiled Figure 02 — a significant upgrade over the first generation. The new robot features improved hands with greater dexterity, onboard AI processing (reducing reliance on remote computing), better cameras and sensors, and a more refined overall design.

The hands in particular represent a step forward. Figure 02's hands have 16 degrees of freedom — meaning 16 independently controllable joints per hand — which enables manipulation of a much wider range of objects and tools. For industrial work, dexterous hands are often the limiting factor: a robot that can walk perfectly but can't handle the parts it needs to move isn't useful.

Figure also announced that Figure 02 would be the version deployed at BMW, with production units beginning to arrive at the Spartanburg facility.

Who's backing Figure and why it matters

The investor list for Figure's $675 million round reads like a who's who of people with a direct stake in humanoid robotics succeeding:

• Microsoft — deeply invested in AI infrastructure and enterprise automation
• OpenAI — whose models power Figure's conversational AI; a strategic partnership as much as an investment
• Nvidia — the dominant supplier of AI computing hardware; humanoid robots are a major market for their chips
• Jeff Bezos — founder of Amazon, which operates some of the most automated warehouses in the world
• Intel Capital — Intel's investment arm, with obvious interest in the computing demands of autonomous robots

These aren't passive financial investors making a bet on a trend. Each of these backers has a strategic reason to want Figure to succeed — and resources to help make it happen. That kind of aligned, strategic investment is meaningfully different from a standard venture round.

When Microsoft, OpenAI, Nvidia, and Jeff Bezos all back the same robotics startup, it's worth asking not just whether they think it will succeed — but why each of them needs it to.

The competitive landscape

Figure isn't the only company in this race. Tesla's Optimus program is arguably the most watched competitor, with Elon Musk making bold claims about eventually producing millions of Optimus robots. Agility Robotics (backed by Amazon) is deploying its Digit robot in Amazon warehouses. 1X Technologies, Apptronik, and a dozen other startups are all working on similar problems.

What distinguishes Figure's position is the combination of serious funding, a real commercial deployment, and the OpenAI partnership. Most competitors have one or two of these. Figure has all three.

The risk is that humanoid robotics is genuinely hard, timelines consistently slip, and the gap between impressive demo and reliable industrial tool has historically been much larger than it looks. Every company in this space is navigating that gap.

What to watch for

The most important near-term signal for Figure will be the BMW deployment. Demo videos are one thing — a robot working reliably in a real manufacturing environment, day after day, is something else entirely. If Figure can demonstrate that Figure 02 operates reliably at Spartanburg, it validates the core thesis and opens the door to much larger commercial agreements.

The second thing to watch is the pace of iteration. Figure has moved from founding to commercial deployment in under three years, with two generations of hardware. If they maintain that pace, Figure 03 is not far away — and each generation has been a meaningful step forward.

The third is pricing. Right now Figure robots are sold into enterprise contracts with pricing that reflects the current stage of the technology. As production scales, costs will fall. The point at which a Figure robot becomes cost-competitive with human labour for a wide range of tasks is the point at which the market becomes very large very quickly.

The honest assessment

Figure AI is one of the most serious players in humanoid robotics — well funded, commercially deployed, and backed by partners with genuine strategic interest in their success. The demos are impressive and the BMW deal is real.

The questions that remain are the same ones facing every humanoid robotics company: Can they achieve the reliability needed for industrial deployment at scale? Can they bring costs down fast enough to expand beyond early-adopter enterprise customers? And can they stay ahead in a race where Tesla, Amazon, and a dozen well-funded startups are all running hard?

None of those questions have definitive answers yet. But Figure has given itself better tools to answer them than almost anyone else in the field.