What are the top Physical AI startups by total funding raised?

Top startups in the Physical AI Market ranked by total funding raised

Here is an updated table that ranks the top startups in the Physical AI Market based on the total amount of funding they have raised to date.

The table also includes the total number of funding rounds, the date and size of the latest round, the financing type (e.g. Series A, equity financing), key investors, the startup’s current status (active, IPO, acquired, or shut down), and a confidence score based on the data collected (we excluded startups with very low data confidence, to make sure everything is reliable).

If you need to dig deeper and get more detailed data, please check our report covering the Physical AI Market.

How far ahead is Waymo in Physical AI funding?

Waymo sits in a different funding class from almost every other Physical AI company.

The $16 billion investment round announced by Waymo in 2026 alone was larger than the cumulative funding of almost every other company in our ranking. The round valued Waymo at $126 billion post-money and included Dragoneer, DST Global, Sequoia Capital and a long list of other institutional investors alongside continued support from Alphabet.

That huge number is easier to understand when we look at what Waymo has become. The company said it completed about 15 million rides during 2025, had passed 20 million lifetime rides when it announced the financing, and was providing more than 400,000 rides per week. Waymo is financing the expansion of an operating transportation network, not just a robotics prototype.

Using the funding figures in our ranking, Waymo alone represents roughly 29% of the cumulative capital raised by all 96 companies. Waymo and Anduril together account for about 41%.

The lead also reflects capital requirements. Operating autonomous fleets across cities requires vehicles, depots, maintenance, mapping and testing infrastructure, regulatory work and years of expensive engineering.

Is Physical AI funding concentrated in just a few companies?

Yes. Physical AI funding is extremely concentrated at the top of the market.

The 96 companies in our working ranking account for roughly $94.7 billion in cumulative funding. The top five alone represent about $46.7 billion, or almost half of that total. The top ten reach approximately $57.5 billion, equivalent to about 61%.

Move further down the ranking and the curve gets even steeper. The top 20 companies represent roughly three-quarters of all funding in the dataset, while the median company has raised only around $287 million.

A handful of extremely capital-intensive companies pull the average upward. Most Physical AI startups are operating on a very different funding scale.

Ranking group	Approx. cumulative funding	Share of dataset funding
Waymo alone	$27.3B	29%
Top 2	$38.6B	41%
Top 5	$46.7B	49%
Top 10	$57.5B	61%
Top 20	$70.8B	75%
Top 50	$86.8B	92%
All 96 companies	$94.7B	100%

What are the biggest Physical AI funding rounds right now?

The biggest Physical AI rounds now regularly reach $1 billion, and several recent financings would have looked extraordinary in robotics only a few years ago.

Waymo's $16 billion raise is the obvious outlier. Anduril followed with a $5 billion Series H at a $61 billion valuation. Shield AI announced a $2 billion financing package consisting of $1.5 billion in Series G equity and $500 million of preferred equity, while Saronic closed a $1.75 billion Series D.

Robotics intelligence companies are now entering the same territory. Skild AI announced a $1.4 billion Series C at a valuation above $14 billion. NEURA Robotics announced a Series C of up to $1.4 billion. Quantum Systems raised $1.2 billion. Wayve closed a $1.2 billion Series D as part of $1.5 billion of capital secured for deployment.

Physical Intelligence's latest Series C was roughly $1.05 billion, while XPENG's robotics business raised more than $900 million.

In our ranking, at least 15 companies show 2026 financing events of $500 million or more. Together, those rounds account for more than $30 billion before counting many smaller transactions.

Are autonomous-driving companies still taking most of the Physical AI money?

Yes. Autonomous driving still absorbs more cumulative capital than any other part of Physical AI.

Using broad categories based on what each company actually builds, autonomous-driving companies in our dataset account for roughly $39 billion of cumulative funding. Waymo explains a large part of that number, but the category also contains Wayve, Motional, Nuro, DiDi Autonomous Driving, Waabi, Applied Intuition and Oxa.

That lead comes from history as much as current enthusiasm. Autonomous driving started attracting billions years before "Physical AI" became a popular label. Companies had to finance sensor systems, vehicle fleets, simulation, data collection and prolonged development cycles long before today's humanoid boom.

The category is still attracting fresh capital. Wayve raised a $1.2 billion Series D and is working with automakers including Mercedes-Benz, Nissan and Stellantis. Waabi closed a $750 million Series C and expanded its original autonomous-trucking strategy into robotaxis.

Humanoids may dominate the conversation right now, but accumulated investor dollars still tell a different story. Self-driving technology remains the largest reservoir of private Physical AI capital.

Is defense becoming Physical AI's second major funding market?

Yes. Defense autonomy has become one of the strongest magnets for Physical AI capital.

Our dataset contains roughly $24 billion of cumulative financing across companies whose main businesses revolve around autonomous military aircraft, drones, vessels, vehicles, counter-drone systems and related defense platforms.

Anduril dominates the category, but the funding is spreading. Shield AI's latest financing package reached $2 billion. Saronic raised $1.75 billion to expand autonomous maritime systems. Quantum Systems closed a $1.2 billion Series D. Mach Industries added another $600 million, while TEKEVER recently announced a $580 million Series D at a $6.4 billion valuation.

Investors are backing autonomy across air, sea and land while military customers are spending more on systems that can operate with fewer humans directly controlling them.

Defense startups can also move toward large contracts while their technology is still evolving. Consumer humanoid companies may eventually address enormous markets, but defense companies can sometimes find billion-dollar demand much earlier through government procurement.

Are humanoid robot startups really raising billions now?

Yes. Humanoid robotics has moved from a niche venture category into a genuine billion-dollar funding race.

Figure has raised more than $1 billion through its Series C alone and previously raised $675 million in Series B financing. NEURA Robotics announced up to $1.4 billion for its Series C. Apptronik says its enlarged Series A has reached more than $935 million after a $520 million extension. XPENG Robotics raised more than $900 million in one private financing.

China is adding another layer of competition. AI² Robotics, Galbot, TARS, LimX Dynamics, Booster Robotics and several others have raised unusually large rounds relative to how young their businesses are.

The speed is the eye-catching part. TARS, for example, raised $455 million in a pre-Series A round roughly a year after the company was founded. Walden Robotics emerged from stealth with a $300 million seed round. Those are amounts that used to belong to mature growth-stage technology companies.

A $500 million round still does not prove that humanoid robots have found a mass market. It shows that investors are willing to finance the manufacturing, data collection and model training needed to find out.

Which humanoid robot startups have raised the most money?

Figure, NEURA Robotics, Unitree, Apptronik and AI² Robotics currently form the best-funded group of humanoid-focused companies in our ranking.

Their financing strategies look very different. Figure has attracted large U.S. technology and financial investors around a vertically integrated humanoid platform. NEURA has built a European investor base that includes technology and industrial partners. Apptronik has Google, Mercedes-Benz, John Deere and QIA among its backers.

China's model is broader. Unitree reached the public market through a Shanghai IPO, while XPENG raised capital directly into a robotics business that sits beside its automotive operations. AI² Robotics, Galbot and several other Chinese startups have mixed private institutional capital with state-linked funds and strategic investors.

The amounts already put humanoid robotics firmly inside the billion-dollar funding economy.

Company	Approx. capital in ranking	Recent financing milestone
Figure	$1.7B+	More than $1B Series C committed
NEURA Robotics	$1.7B	Series C of up to $1.4B
Unitree Robotics	$1.1B+	Roughly $904M raised in IPO
Apptronik	nearly $1B	$520M Series A extension
AI² Robotics	about $944M	$739M Series C
XPENG Robotics	$900M+	More than $900M private financing
Galbot	about $887M	Roughly $362M recent round
RobotEra	about $731M	Multiple Chinese financing rounds
TARS	about $697M	$455M pre-Series A

Is Skild AI becoming one of the best-funded robotics startups?

Yes. Skild AI has gone from a young robotics-model company to one of Physical AI's largest private financing stories.

The company raised a $300 million Series A in 2024. Its latest Series C was $1.4 billion and valued Skild above $14 billion. Depending on how intermediate rounds and tranches are reconciled, public funding histories now place cumulative capital at roughly $2 billion or more.

Skild's pitch is different from that of a humanoid manufacturer. It wants to build a general-purpose intelligence layer capable of controlling different kinds of robots instead of betting everything on one robot body.

That resembles the model-layer strategy investors already understand from generative AI. If one intelligence system can work across mobile robots, industrial arms, quadrupeds and humanoids, the potential market becomes much larger than sales of a single machine.

Commercial evidence is starting to appear too. Skild said recently that it had crossed $100 million in annual recurring revenue within ten months of its first commercial deployment. If that develops into durable revenue, Skild will have something many heavily financed robotics startups still lack: a sizable software-like commercial base.

Who has raised more money, Skild AI or Physical Intelligence?

Skild AI and Physical Intelligence are now in roughly the same funding tier, with both around the $2 billion mark depending on how individual rounds are reconciled.

Physical Intelligence's financing history is particularly striking because the company was only founded in 2024. A roughly $70 million seed financing was followed by about $400 million, then roughly $600 million, and then a $1.05 billion Series C. That puts cumulative disclosed financing close to $2.1 billion.

Skild has followed a similarly steep path, including its $300 million Series A and $1.4 billion Series C.

The two companies are chasing a similar prize: general robotic intelligence that can move across tasks and hardware. Skild is publicly talking more about deployments, partnerships and recurring revenue. Physical Intelligence has historically been more research-driven, with its founders willing to spend heavily on models and compute before setting a fixed commercialization timetable.

Who has raised more funding, Figure or Apptronik?

Figure remains better funded overall, but Apptronik has closed the gap much faster than it looked capable of doing a year ago.

Figure's Series C exceeded $1 billion at a $39 billion post-money valuation, following its earlier $675 million Series B and earlier financing. Apptronik's initial $415 million Series A was already large, then the company added another $520 million through a Series A extension, bringing that single financing program above $935 million.

The investor lists also show where each company expects demand to come from. Figure has attracted NVIDIA, Brookfield, Intel Capital, Salesforce and other technology and financial investors. Apptronik's list contains Google, Mercedes-Benz, John Deere, AT&T Ventures and Qatar Investment Authority.

Apptronik has something unusually useful for a humanoid startup: several investors that could also become technology partners, industrial users or distribution channels.

At this point, the funding gap matters less than execution: manufacturing robots reliably, deploying them for long periods and proving that customers will pay for useful work rather than demonstrations.

Is China catching the United States in Physical AI funding?

China is catching up fastest in humanoids and embodied AI, although the United States still holds the biggest individual funding positions.

The U.S. ranking is boosted enormously by Waymo, Anduril, Shield AI, Physical Intelligence, Skild AI, Figure, Waabi, Apptronik and other heavily funded companies. No Chinese startup currently approaches Waymo's cumulative funding.

The picture changes when we isolate newer humanoid and embodied-AI companies. XPENG Robotics raised more than $900 million in a single financing. AI² Robotics reportedly raised about $739 million in its Series C. Galbot added roughly $362 million. TARS raised $455 million in a pre-Series A round. Several companies that barely existed two years ago are already financed at levels normally associated with mature growth companies.

China also benefits from a dense manufacturing supply chain around motors, actuators, batteries, sensors and industrial automation. Equivalent venture dollars can therefore be deployed inside a different cost structure.

The U.S. still leads on absolute private capital at the very top. In the current generation of humanoid companies, the gap is much tighter.

How important is XPENG Robotics' $900 million funding round?

XPENG Robotics' financing is one of the clearest signs that China's humanoid race has moved into large-scale industrial financing.

XPENG announced that its robotics business had raised more than $900 million at a post-money valuation above $6.3 billion. IDG Capital led the financing, with Gaorong Ventures participating and Tencent and Alibaba providing strategic backing.

The round was unusually large even by the new standards of humanoid robotics. XPENG described it as the largest single private capital raise in China's embodied-AI industry at the time.

Its importance goes beyond the number. XPENG already operates a sophisticated automotive engineering and manufacturing organization. A robotics division connected to that ecosystem can reuse knowledge in batteries, motors, embedded computing, AI, supply chains and mass manufacturing.

It does not guarantee XPENG will win in humanoids. It does give the company a very different industrial base from a typical robotics startup.

Are giant seed and Series A rounds becoming normal in Physical AI?

They are not normal across startups generally, but they are becoming surprisingly common inside Physical AI.

Walden Robotics launched publicly with a $300 million seed round led by Toyota and Deviation Capital. NVIDIA, Boeing, Samsung Ventures, Prologis Ventures and CoreWeave Ventures also participated.

Rhoda AI emerged from stealth with a $450 million Series A. TARS raised $455 million before reaching a conventional Series A. Apptronik's Series A program eventually grew beyond $935 million.

Physical AI is compressing the normal venture-funding timeline. Investors are putting growth-stage amounts of capital into companies that are still very young.

There is a practical reason. A software startup can often test demand with a small engineering team and cloud infrastructure. Physical AI companies may need robot hardware, custom components, factories, training data, simulation, large model-training clusters and deployment teams before they can prove the business.

The risk is just as obvious: huge early rounds can price in commercial success long before the company has demonstrated it.

Why are automakers investing so much in Physical AI startups?

Automakers are becoming major Physical AI investors because robotics and autonomous driving increasingly overlap with capabilities car companies already need.

Mercedes-Benz has backed both Wayve and Apptronik. Toyota co-led Walden Robotics' $300 million seed financing. Volkswagen-linked capital participated in Mind Robotics, a company spun out of Rivian's ecosystem. Nissan and Stellantis have invested in or partnered with Wayve.

These companies are buying more than financial exposure. Automakers have factories where industrial and humanoid robots can be tested, vehicle platforms where autonomous-driving software can be deployed and enormous procurement organizations that can turn a successful pilot into a large contract.

Their factories also provide something robotics startups badly need: real-world data.

A robot moving parts around a factory, an autonomous vehicle navigating a city and a foundation model controlling machines all depend on perception, planning, simulation and reliable physical execution. Automakers increasingly treat those capabilities as part of the same strategic technology stack.

Why does NVIDIA keep appearing in Physical AI funding rounds?

NVIDIA is backing Physical AI companies because every successful robot can eventually create demand for more training, simulation and edge-computing infrastructure.

Its investments and partnerships stretch across robot intelligence, humanoids, autonomous vehicles and industrial robotics. Skild AI received backing from NVentures. NVIDIA participated in Figure's funding, appears among NEURA's backers, invested in Walden Robotics and has relationships with companies including Wayve, Nuro and FieldAI.

The financial investment is only one part of the strategy. NVIDIA is simultaneously building the infrastructure those companies can use: Isaac robotics software, Cosmos world models, simulation tools, GR00T models and edge hardware.

NVIDIA does not need to know whether humanoids, robotaxis, industrial arms or autonomous construction equipment becomes the largest Physical AI market. It can supply computing infrastructure to several of them.

Who are the biggest recurring investors in Physical AI startups?

NVIDIA, SoftBank, Sequoia, Andreessen Horowitz, General Catalyst, Lightspeed and several large strategic investors repeatedly appear across major Physical AI financings.

Their strategies are not identical. SoftBank has made particularly aggressive bets on robot intelligence, including Skild AI. Sequoia appears across autonomous systems and robotics. Andreessen Horowitz has exposure to defense, robotics and autonomy. Lightspeed has backed companies including Skild AI, Physical Intelligence and Reliable Robotics.

Industrial investors form another group. Google has backed Apptronik. Mercedes-Benz appears around both autonomy and humanoids. Toyota is funding general-purpose robotics. Qualcomm, Arm and AMD invested directly in Wayve. Salesforce Ventures appears in several industrial AI and robotics financings.

Sovereign and state-linked capital matters too. QIA backed Apptronik. European institutions are present in NEURA's financing. Chinese state and municipal funds appear repeatedly across embodied-AI companies.

There is no single "robotics investor club" here. AI investors, industrial companies, defense investors and sovereign capital are converging on the same physical technologies.

Are robot foundation-model startups raising as much money as robot manufacturers?

They are getting surprisingly close, and the funding gap between software intelligence and robot hardware has almost disappeared.

Physical Intelligence has raised around $2.1 billion. Skild AI sits around the same general level depending on how its rounds are reconciled. Generalist AI, FieldAI, Dyna Robotics and other companies are also raising substantial amounts around the idea that the intelligence layer itself can become the valuable part of robotics.

The strongest argument for this strategy is reuse. A humanoid manufacturer has to make its particular robot commercially successful. A robot foundation-model company can theoretically sell intelligence across arms, mobile robots, humanoids, warehouse systems and machines that have not even been designed yet.

Investors already understand that economic model from software and generative AI.

Robot intelligence still cannot stay completely detached from hardware. These models need huge quantities of physical interaction data, and they need access to machines to generate it. That helps explain why Skild acquired Zebra Technologies' robotics business and why foundation-model companies are increasingly building deployment relationships.

Are traditional industrial robotics startups being left behind?

Traditional industrial robotics companies are raising less money than frontier Physical AI labs, but commercially focused automation remains one of the healthier parts of the market.

Standard Bots raised $200 million at a $1 billion valuation and says it is expanding U.S. manufacturing capacity. Dexterity has raised hundreds of millions for warehouse robotics. Path Robotics has attracted substantial funding around autonomous welding. Mujin, Quicktron, RobCo, Nimble, Mytra and AMP Robotics are all well funded relative to most industrial startups.

These businesses usually attack specific economic problems. Welding, pallet movement, sorting, machine tending and warehouse fulfillment already have customers with measurable labor costs.

That can make them less spectacular fundraising stories but easier businesses to evaluate.

Frontier humanoid and robot-brain companies promise broader eventual markets, while industrial robotics firms often start with a simpler purchasing question: does this machine perform the task reliably enough to save money?

Which Physical AI startups have raised funding most recently?

Recent fundraising remains unusually active across defense, humanoids, robot intelligence and industrial automation.

TEKEVER announced a $580 million Series D at a $6.4 billion valuation. Mach Industries added $600 million to its Series C financing. XPENG Robotics raised more than $900 million. Walden Robotics launched with $300 million. Quantum Systems raised $1.2 billion.

Several smaller transactions matter too because they show capital spreading beyond the headline companies. Standard Bots raised $200 million. THEKER Robotics raised $85 million. WIRobotics raised $68 million. RoboForce added $52 million.

There is also capital waiting behind deals that have not formally closed. FieldAI has recently been reported to be seeking roughly $700 million at around a $10 billion valuation. We would not count that money in cumulative funding until the financing is actually completed.

The fundraising boom is not confined to one robot type. Capital is moving at the same time into autonomous vehicles, defense platforms, humanoids, industrial machines and general-purpose robot brains.

Company	Recent round or financing	Main category
Quantum Systems	$1.2B	Defense autonomy
XPENG Robotics	$900M+	Humanoids
Mach Industries	$600M	Defense
TEKEVER	$580M	Autonomous defense aircraft
Walden Robotics	$300M	General-purpose robotics
Standard Bots	$200M	Industrial robotics
BRINC	$125M	Emergency-response drones
THEKER Robotics	$85M	Industrial robotics

Does raising more money mean a Physical AI startup is winning?

No. Funding tells us who investors are willing to finance, but it does not tell us by itself who will build the best business.

Waymo is the easiest example of funding and operating scale moving together. Its enormous financing sits alongside millions of commercial rides and a growing service footprint.

Other companies are much earlier. A foundation-model startup can raise more than $1 billion before we know whether customers will eventually pay enough for its robot intelligence. A humanoid manufacturer can finance a factory before anyone has proved how many humanoids businesses will buy at sustainable prices.

Physical AI requires unusually large upfront spending, so big funding totals are not automatically irrational. They just should not be treated as product rankings.

Commercial leadership still needs separate evidence: revenue, production volume, deployment hours, customer renewals, unit economics, reliability and actual utilization.

Is Physical AI funding still accelerating right now?

Yes. Physical AI funding is still accelerating, even after several years of heavy investment in autonomous vehicles and robotics.

Crunchbase estimated that companies in its broader Physical AI definition raised about $47.4 billion across 521 deals during the first half of 2026, almost four times the roughly $12 billion raised during the second half of 2025.

Different market definitions produce different totals, so we would not merge that number directly with our narrower startup ranking. The direction is much harder to dispute.

Our own company-level evidence points the same way. Billion-dollar financings have appeared in autonomous driving, defense, humanoids and robot foundation models during the same period. Early-stage companies are raising hundreds of millions. Strategic investors such as automakers, semiconductor companies and industrial groups are also joining rounds that might previously have been dominated by venture firms.

Physical AI investment is no longer dependent on robotaxis alone. Several categories are now capable of generating very large rounds at the same time.

What types of Physical AI startups attract the most funding?

Autonomous driving still leads by cumulative funding, defense comes next, and humanoid robotics has become the fastest-moving challenger.

When we group the 96 companies in our ranking by their main activity, autonomous-driving businesses account for about $39.4 billion. Defense autonomy represents roughly $24 billion. Humanoid-focused companies contribute around $12.5 billion, while robot foundation-model businesses add almost $7 billion.

These numbers should be read as directional rather than as official market-size statistics because companies increasingly overlap categories. Wayve describes itself as embodied AI but sells autonomous-driving software. NEURA combines humanoids with a broader Physical AI platform. Skild wants its models to control many robot forms.

Still, the hierarchy is useful. Investors have so far placed the most money where Physical AI already connects to huge existing budgets: transportation and defense. Humanoids and general robotics are now large enough to challenge that historical pattern.

Physical AI category	Approx. cumulative funding in ranking	What drives the capital
Autonomous driving	$39.4B	Robotaxis, trucking, vehicle software
Defense autonomy	$24.0B	Drones, aircraft, vessels, military systems
Humanoid robotics	$12.5B	Manufacturing, logistics, general labor
Robot foundation models	$6.9B	General-purpose robot intelligence
Autonomous delivery	$4.0B	Drones and delivery vehicles
Industrial robotics	$3.4B	Factory and warehouse automation

Why are Physical AI startups raising so much more money than normal software startups?

Physical AI consumes capital in several places at once: AI models, hardware, manufacturing, data and real-world operations.

A software startup can often deploy a new version of its product to thousands of customers without producing anything physical. A robotics company cannot.

Robots need motors, actuators, batteries, sensors, chips and mechanical systems. Hardware has to be manufactured, shipped, repaired and redesigned. Models need real-world data, which can require fleets of machines and human operators. Simulation and training consume expensive computing resources.

Then comes deployment. Putting a robot into a warehouse, factory, public road or military environment creates integration and safety work that does not exist for ordinary cloud software.

That is also why funding rounds increasingly combine traditional venture investors with companies such as NVIDIA, Google, Toyota, Mercedes-Benz, Qualcomm, John Deere and Boeing. The strategic investor may bring manufacturing expertise, computing platforms, factories or customers alongside money.

Is there already too much money chasing Physical AI startups?

Some Physical AI valuations clearly assume a huge amount of future success, but the market is still too early to call the entire funding wave irrational.

There are reasons to be cautious. Companies only a few years old are raising hundreds of millions before establishing mature commercial models. Several robot-intelligence companies are worth billions despite limited public revenue disclosure. Humanoid manufacturers are financing large production ambitions before anyone knows how quickly customers will deploy thousands of machines.

That creates plenty of room for failures.

At the same time, the businesses do not all have the same risk profile. Waymo has commercial rides. Defense companies are winning contracts. Industrial robotics firms sell machines into factories. Robot-model companies are beginning to report meaningful commercial deployments.

The useful distinction is between funding the infrastructure required to reach scale and funding a valuation that assumes scale has already been achieved. We are seeing both.

What does the Physical AI funding ranking actually tell us about who is winning?

The ranking shows that Physical AI has already become a major capital market, but funding leadership currently belongs to autonomy and defense rather than humanoid robots.

Waymo is the clear funding outlier. Anduril has built a second enormous capital pool around autonomous defense. Behind them, the market becomes far more competitive, with autonomous-driving companies, robot foundation-model developers and humanoid manufacturers increasingly occupying the same billion-dollar funding tier.

The biggest change is happening below the two leaders. Skild AI, Physical Intelligence, Figure, NEURA Robotics, Apptronik, XPENG Robotics, Quantum Systems, Saronic and a growing group of Chinese embodied-AI companies can now raise amounts that were almost unthinkable for robotics startups a few years ago.

Investors are financing several competing architectures at once. The next phase will be about which companies can convert those billions into machines that customers use repeatedly, at scale, for economically useful work.