Figure AI

What Figure AI is

Figure AI builds general-purpose humanoid robots: machines with a human-like body plan, meant to do physical work in workplaces that were built for people. The company is based in Sunnyvale, California, and its founder and CEO is Brett Adcock.^[3][4]

The bet behind the humanoid form is a practical one. Factories, warehouses and stores are already shaped around human bodies — benches at waist height, stairs, door handles, tools sized for hands. A robot that matches those dimensions can be put into an existing workplace instead of requiring the workplace to be rebuilt around it. The hard part is that balance, grasping and perception have to work together, reliably, for a whole shift, which is why general-purpose humanoids remain much rarer than fixed industrial arms.

BMW and robots-as-a-service

In January 2024, Figure signed a commercial agreement with BMW Manufacturing to deploy general-purpose humanoid robots in automotive manufacturing.^[3] Car assembly is a demanding first market: tasks repeat but vary, cycle times are tight, and a stopped line is expensive.

TechCrunch reported that Figure was effectively leasing its systems through a robotics-as-a-service model.^[3] Rather than selling a robot outright, the supplier keeps the machine and charges for its use. That lowers the barrier for a customer trying an unproven technology, and it turns a one-time sale into recurring revenue — but it also leaves the capital cost, maintenance and upgrades with Figure.

Two funding rounds

A 2024 report said Figure had raised US$675 million in a round backed by NVIDIA, OpenAI, Microsoft and Jeff Bezos.^[4]

On 16 September 2025 the company announced more than US$1 billion of committed capital in its Series C, at a US$39 billion post-money valuation.^[1] The round was led by Parkway Venture Capital, with Brookfield Asset Management, NVIDIA, Macquarie Capital, Intel Capital, Align Ventures, Tamarack Global, LG Technology Ventures, Salesforce, T-Mobile Ventures and Qualcomm Ventures participating.^[1]

The list of participants is worth reading closely. It mixes chipmakers (NVIDIA, Intel Capital, Qualcomm Ventures), asset managers with large industrial and infrastructure holdings (Brookfield, Macquarie), a telecom operator (T-Mobile Ventures), an enterprise software company (Salesforce), a consumer electronics group (LG Technology Ventures) and a venture firm leading the round. Each brings a different reason to care — compute, factories, connectivity, software integration, components — which suggests humanoid robots are being treated as a platform that many industries will touch rather than as a single product.

What the money is meant to do

Figure said the funding would scale its Helix AI platform and BotQ manufacturing, expand real-world deployments, build next-generation GPU infrastructure and launch advanced data collection.^[1]

Those items describe a loop rather than a shopping list. Helix is the company's AI platform; BotQ is its manufacturing operation; deployments put robots into real workplaces; data collection captures what those robots encounter; and GPU infrastructure is the computing needed to turn that data into better behaviour. Working robots generate data, data improves the models, better models make the robots more useful — the standard argument for why this category attracts capital before it has large revenue.

On production, Figure said BotQ's first-generation manufacturing line would be capable of producing up to 12,000 humanoid robots per year, with a goal of 100,000 robots over the following four years.^[2] Humanoid robot output has so far been measured in small numbers, so a five-figure annual line implies a different kind of operation, with parts supply and testing as much of a constraint as assembly itself.

Figure 03

Figure introduced its third-generation robot, Figure 03, on 9 October 2025.^[2] The timing matters: a new hardware generation is what a high-volume line is meant to build and what deployment partners are meant to test.

Where it sits

Figure is one of the companies that moved humanoid robots from demonstration videos into commercial contracts and large funding rounds. Its position rests on three things that have to hold at once: a robot reliable enough to be leased rather than demonstrated, a manufacturing line able to supply fleets rather than units, and an AI platform that keeps improving from the data those fleets produce. The BMW agreement, the BotQ capacity figures and the Series C each address one of those, and each is a claim that still has to be proven in operation.