In 2025, humanoid robotics attracted over $10.3 billion in venture funding across more than 450 individual rounds while shipping roughly 13,000 actual units — a ratio of investment to deployed hardware that defines this market more precisely than any individual funding headline does.
After going through the deal data, the investor lists, and the valuation multiples that sit behind the biggest rounds, the investment picture in 2026 is one of extraordinary capital concentration: the top five deals account for roughly 75% of all disclosed funding, megarounds have become the norm rather than the exception, and the investors writing the largest checks are explicitly betting on a $200 billion market by 2035 that doesn't fully exist yet.
This is the latest piece in our ongoing humanoid robot series — previous entries covered enterprise deployments, home robots, Chinese manufacturers, labor market impacts, and military applications. This piece focuses entirely on the money: who's funding whom, which investors keep showing up, and what the deal structures reveal about where the smart money thinks this market is actually heading.
The Investment Landscape at a Glance
| Company | Total Funding | Valuation | Key Investors | What They're Betting On |
|---|---|---|---|---|
| Figure AI | $1.75B | $39B | OpenAI, Microsoft, Nvidia, Amazon, Intel, LG, Samsung | Premium humanoid hardware + in-house Helix AI |
| Skild AI | $2.2B | $14.3B | Sequoia, Coatue, Amazon, Bezos Expeditions | Foundation model AI that runs across any robot hardware |
| Physical Intelligence (π) | ~$1B+ | ~$11B | OpenAI, Alphabet, Jeff Bezos, Thrive Capital, Sequoia | Hardware-agnostic "Android for robots" AI platform |
| NEURA Robotics | $1.48B | ~$7B | Tether, Qualcomm, Amazon, Nvidia, Bosch, Schaeffler, EIB | Cognitive robots for industrial and daily settings (Europe) |
| Apptronik | $935M+ | $5.5B+ | Google, B Capital, Mercedes-Benz, John Deere, QIA | Apollo humanoid; exclusive Google DeepMind hardware partner |
| Agility Robotics | ~$180M (disclosed) | $2.12B | CapitalG, Jeff Bezos, Thrive Capital, Sequoia | Only humanoid generating real commercial revenue (Digit/GXO) |
The Concentration Problem: Five Deals, 75% of the Capital
The single most important structural fact about humanoid robotics funding in 2025–2026 is the concentration. New Market Pitch's analysis of every disclosed equity round between May 2025 and April 2026 found that 10 of 16 disclosed deals cleared $50 million, and 8 cleared $100 million — but stripping the top five deals removes 75.2% of all disclosed capital, leaving just $717.9 million across the remaining eleven rounds. This is an unusual pattern even for a capital-intensive hardware market: it means the handful of companies at the top are being priced like category leaders, while everyone else is getting funded at a substantial discount regardless of technical merit.
The rule that emerges from the data is specific: investors pay up for breadth plus scale path, not narrow technical excellence alone. Companies that combine hardware, autonomy, deployment narrative, and manufacturing ambition get priced like category leaders. Specialists get funded at a clear discount. That explains why Figure AI, with only around 150 commercial units shipped at the time of its $39 billion valuation, and Physical Intelligence, with no disclosed revenue at its $11 billion valuation, command multiples that would be considered extraordinary in any other hardware sector. The bet isn't on current revenue — it's on platform positioning before the market reaches $200 billion.
Who's Writing the Biggest Checks and Why
The investor list for humanoid robotics in 2026 reads like a reunion of the biggest names in tech, with a few surprising additions. Jeff Bezos has backed Agility Robotics and Physical Intelligence personally, separate from Amazon's corporate investments. Amazon itself is a strategic investor in NEURA Robotics and Agility simultaneously — hedging across hardware and AI software platforms rather than betting on a single robot. Google has designated Apptronik as the exclusive hardware partner for its Gemini Robotics project while separately backing Physical Intelligence through CapitalG and Alphabet.
Nvidia shows up in multiple humanoid investments — Figure AI, NEURA Robotics, and others — a logical position for a company whose chips power the training and inference stacks of every humanoid AI. Nvidia's investment thesis is essentially platform: whoever wins the humanoid market will need massive amounts of GPU compute, so having equity stakes across the leading platforms is a bet on the infrastructure layer rather than any specific robot design. NEURA's June 2026 Series C also brought in Qualcomm, Bosch, Schaeffler, and the European Investment Bank — a notably different investor mix than the U.S.-heavy rounds that dominate the sector, and a signal that European industrial heavyweights see humanoid robots as a strategic supply chain issue, not just a venture bet.
On the China side, the government is explicitly in the game. China's 2026 five-year plan includes over $20 billion in subsidies to the robotics industry through loans, tax credits, and state-backed venture capital. China's Ministry of Industry and Information Technology reported a 33.2% year-on-year increase in industrial robot output in Q1 2026, following 610 investment deals valued at $7.9 billion by end-2025. This is state-directed industrial policy operating at a scale that U.S. and European venture capital can't match deal-for-deal — and it's a meaningful part of why Chinese humanoid companies can ship at lower cost and higher volume than Western counterparts.
The AI Software Layer: Where the Real Valuation Premium Lives
The most revealing thing about the 2026 humanoid investment landscape isn't which robot companies are most funded — it's that two of the highest-valued companies (Skild AI at $14.3 billion and Physical Intelligence at $11 billion) don't build robots at all. Both build AI software platforms designed to run across any robot hardware, and both have attracted checks from investors who see the pattern from the smartphone era playing out again: just as Android and iOS captured more value than most handset manufacturers, a foundation model layer for robotics might ultimately be worth more than the robots themselves.
Skild AI's $2.2 billion in total funding — more than any hardware-first humanoid company except Figure — comes from Sequoia, Coatue, Amazon, and Bezos Expeditions betting on exactly that thesis. Physical Intelligence's Π0 model, designed as a general-use AI platform that can run across different robot bodies rather than training each robot for specific tasks, has attracted backing from OpenAI, Alphabet, and Jeff Bezos personally. The thesis in both cases is the same: whoever builds the AI layer that makes any robot more capable is better positioned than whoever builds the best robot body for one specific task.
Where the Honest Tension Lives
The clearest honest tension in humanoid investment right now is between the market projections driving valuations and the deployment reality those valuations are based on. The market today is $2 to $3 billion by revenue — small enough that a single large enterprise contract moves the needle. Barclays' base case puts it at $35 billion by 2035; the high case is $200 billion. Morgan Stanley has published a $5 trillion estimate for 2050. These projections are the foundation of valuations like Figure's $39 billion, which implies roughly thirteen times the current entire market cap of humanoid robotics as a sector.
SoftBank CEO Masayoshi Son, who has called physical AI and robotics "where the next trillion-dollar company will emerge," and Wedbush's Dan Ives, who called humanoid robots "one of the biggest market opportunities in the AI Revolution," are betting that the gap between those projections and the current reality will close faster than most timelines assume. The investors who are more skeptical point to a consistent pattern in robotics: the market has been "two to three years away" from mainstream deployment for over a decade, and the specific challenge of general-purpose dexterity in unstructured environments hasn't been solved despite enormous compute and capital.
Both positions are defensible with the available data. What isn't defensible is treating the current funding pace as proof the deployment reality is already here — 13,000 units shipped globally in 2025 against $10.3 billion invested is a ratio that would be considered speculative in any other sector. The investors writing the biggest checks know this and are betting anyway, which tells you something specific about their conviction that the platform race needs to be won before the revenue arrives.
Frequently Asked Questions
How much money has been invested in humanoid robotics?
Global humanoid robotics attracted over $10.3 billion in venture funding across more than 450 individual rounds in 2025 alone, with robotics companies broadly raising $55.8 billion in the first half of 2026 according to Dealroom — though the humanoid sector specifically remains heavily concentrated, with the top five deals accounting for roughly 75% of all disclosed capital.
Which humanoid robot company has raised the most money?
Figure AI leads with $1.75 billion in total funding at a $39 billion valuation as of mid-2026, though Skild AI — which builds AI software for robots rather than robots themselves — has raised $2.2 billion at a $14.3 billion valuation, making it the single highest-funded company in the humanoid AI ecosystem by total capital raised.
Why are investors backing humanoid robots when so few are actually deployed?
The investment thesis is explicitly forward-looking: analysts project the humanoid market will grow from roughly $2 to $3 billion today to $35 to $200 billion by 2035, and investors are betting on platform positioning before that market develops — the same pattern that played out in cloud computing and mobile operating systems, where early platform bets captured disproportionate returns.
What role is China's government playing in humanoid robot funding?
China's 2026 five-year plan includes over $20 billion in subsidies to the robotics industry through loans, tax credits, and state-backed venture capital, and the Ministry of Industry and Information Technology reported 610 investment deals valued at $7.9 billion in humanoid-related sectors by end-2025 — state-directed investment at a scale that Western venture capital cannot match dollar-for-dollar.
What is Physical Intelligence and why is it valued at $11 billion with no revenue?
Physical Intelligence is building AI foundation models designed to run across any robot hardware rather than training each robot for specific tasks — the "Android for robots" thesis — and has attracted backing from OpenAI, Alphabet, Jeff Bezos, Thrive Capital, and Sequoia on the bet that an AI platform layer will ultimately capture more value than any individual robot design, similar to how mobile operating systems outperformed most handset manufacturers.