Robotics Startups 2026: Figure, Physical Intelligence, 1X — Who Has the Real Lead?

Robotics startups raised more than $7 billion across 2024–2025, and the humanoid race now has a clear top tier: Figure AI (reportedly raising near a $39 billion valuation), Physical Intelligence ($2.4 billion), and 1X Technologies. That's the short answer. The longer answer is more interesting.

The interesting part is that “who has the lead” depends entirely on what you measure. Figure leads on capital and on having an actual robot factory. Physical Intelligence leads on the AI model layer that may turn out to matter more than any single robot body. 1X leads on the home, Apptronik leads on industrial partnerships, and Skild leads on a body-agnostic brain. There is no single scoreboard — so this post builds three of them.

Robotics Startups 2026 Funding: Who's Raised the Most and Who Leads

Robotics startups raised more than $7 billion across 2024 and 2025, with humanoid and physical-AI companies taking the largest share. Figure AI leads on reported valuation after pursuing a roughly $1 billion Series C near a $39 billion mark in 2025, followed by Physical Intelligence at $2.4 billion, with 1X, Apptronik, and Skild AI close behind. On capital raised Figure leads; on the AI model layer, Physical Intelligence does.

To make sense of it, separate the two things investors are actually funding. One is a hardware bet: build a physical humanoid that can do real work. The other is a software bet: build the foundation model — the “brain” — that lets any robot generalize across tasks. The companies that have raised the most are split across that line, which is why a single ranking misleads. You can track how these private valuations stack up against the broader market on the AI Valuations dashboard.

The Robotics Startups 2026 Funding Scoreboard

Here is the capital scoreboard for the leading robotics startups as of mid-2026. The numbers blend disclosed rounds with widely reported figures; where a raise was reported but not formally confirmed, it is marked accordingly. Tesla's Optimus is included for context but is not a standalone startup — it sits inside Tesla.

Figures are mid-2026 estimates compiled from company announcements, PitchBook, Crunchbase, and reporting by Bloomberg, The Information, and Reuters. “Reported” valuations reflect rounds that were widely reported but not formally confirmed by the company; “est.” figures are author estimates derived from disclosed round sizes and ownership. Total raised is cumulative across all known rounds.

Figure AI: The Capital Leader With a Factory

Figure AI is the clearest capital leader. After a $675 million Series B in early 2024 at a $2.6 billion valuation — backed by Microsoft, OpenAI, Nvidia, Intel Capital, and Jeff Bezos — the company reportedly pursued a Series C near $1 billion at a roughly $39 billion valuation in 2025. That is a 15x markup in under two years, the kind of repricing that only happens when investors believe a company can define a category.

What Figure actually has to show for it is more concrete than most. It announced BotQ, a high-volume manufacturing facility targeting up to 12,000 humanoids per year initially and a stated path toward 100,000 units over four years. It shipped the Figure 02 and then Figure 03 robots, and it moved its control software in-house with Helix, a vision-language-action model, after parting ways with an earlier OpenAI collaboration. Figure's thesis is vertical integration: own the robot, the factory, and the brain.

The risk is equally clear. A $39 billion valuation implies enormous future revenue from a product still largely in pilots. Owning the full stack is capital-intensive and slow, and any humanoid company is one safety incident or missed factory ramp away from a brutal down round. Figure leads the funding race, but it has also raised the bar it must clear.

Physical Intelligence and Skild: The Model-Layer Bet

Physical Intelligence raised $400 million in late 2024 at a $2.4 billion valuation — remarkable for a company founded that same year, backed by Bezos, OpenAI, Thrive Capital, and Lux Capital. Its bet is the opposite of Figure's: rather than build a humanoid, it builds π0 (“pi-zero”) and successor models — general-purpose robot foundation models designed to control many different robot bodies. If the brain is the durable asset, Physical Intelligence wants to be the OpenAI of robots, not the Tesla.

Skild AI plays a similar game from a different angle. It raised about $300 million in 2024 at a $1.5 billion valuation from SoftBank, Coatue, and Bezos to build a body-agnostic “robot brain” trained across simulation and real-world data. The model-layer companies are cheaper to fund than full-stack hardware plays and, if they win, capture value across every robot maker rather than competing as one of many.

The catch: a model is only as good as the data and the hardware it runs on. Foundation-model robotics companies depend on partners to provide bodies and real-world deployment data, and the gap between an impressive demo and reliable, generalized performance in unstructured environments is still wide. The model layer may be where the lead ultimately sits — but it is the hardest lead to prove with a video.

1X, Apptronik, and Agility: Picking Their Battlefields

The next tier of robotics startups isn't losing — it's specializing. 1X Technologies, the Norwegian company backed by OpenAI and EQT Ventures, is going after the hardest market of all: the home, with its NEO humanoid. Apptronik raised a $350 million Series A in early 2025 with Google participating, and is targeting industrial and logistics work with its Apollo robot and partnerships with the likes of Mercedes-Benz. Agility Robotics, maker of Digit, has roughly $580 million raised and the most real-world warehouse deployment experience, including pilots with Amazon.

This is the strategic split that defines the category. The home is the largest market but the hardest and least forgiving on safety. Warehouses and factories are smaller but structured, repetitive, and willing to pay for labor automation now. Most of the durable early revenue in robotics will come from these industrial settings — which is also where the overlap with defense and dual-use robotics is growing fastest.

Who Actually Has the Lead in Robotics in 2026?

So who wins? It depends on which scoreboard you trust — and here they are side by side.

Assessments are author analysis as of mid-2026, based on company disclosures, product demonstrations, and reporting by Bloomberg, The Information, and Reuters. “Leader” reflects relative position on each dimension, not a guarantee of future outcome; the category remains early and pre-revenue for most players.

My read: Figure has the best position to win the headline race because capital plus a factory plus an in-house model is the hardest combination to replicate — but Physical Intelligence has the most asymmetric upside, because if the model layer generalizes, it wins across everyone's hardware. The likeliest 2026–2027 outcome isn't one winner; it's a split where a hardware leader and a model leader both emerge, and the most dangerous competitor to all of them is Tesla, which has the one thing no startup has: the balance sheet and supply chain to mass-produce. Watch how the funding rounds and valuations move on the AI Valuations dashboard and the Unicorns tracker.