Physical AI Funding Tracker (360 deals)

Our funding tracker covering the Physical AI Market

Below is the table listing all the deals. You may find our methodology at the end of this page.

And, if you want to gain a deeper understanding of the market and its current dynamics, get our report covering the Physical AI Market.

Is Physical AI really a $63 billion funding market right now?

Physical AI really has attracted about $63 billion in our research window, but an ordinary company in this market operates on a financial scale far below that headline.

We count 359 qualifying financings across 289 companies, with at least $62.96 billion of recorded capital. Eight deals have no usable disclosed amount and 15 are explicit lower bounds, so the true total is somewhat higher.

The useful number for understanding a typical financing is the $28 million median. The average is $179.4 million, more than six times higher, because a relatively small number of huge rounds pull it upward.

Remove every financing above $50 million and only $3.65 billion remains. That means roughly 94% of all the money came through larger rounds, even though most financing events were below that threshold.

So yes, Physical AI is very well funded today. The catch is that the headline combines two very different worlds: hundreds of normal venture rounds and a much smaller group of companies capable of raising hundreds of millions or even billions of dollars.

Metric	Physical AI funding
Qualifying financings	359
Funded companies	289
Recorded capital	$62.96B+
Median disclosed round	$28.0M
Average disclosed round	$179.4M
Average / median	6.4x
Capital left after removing >$50M rounds	$3.65B

How many companies are actually getting funded in Physical AI?

Physical AI currently has a broad funded base of 289 companies, although most of them have only raised once during our research window.

We find 237 companies with one qualifying financing, 40 with two and just 12 with three or more.

In other words, 82% of funded Physical AI companies appear only once.

That changes the way we should read the 359-deal total. There is plenty of company creation and early financing, but repeated access to capital is much rarer.

The minority that does come back to market is economically much more important. Companies with more than one qualifying financing collected about $45.09 billion, compared with $17.87 billion for companies that appeared only once.

Just 18% of companies therefore captured roughly 72% of the money by raising repeatedly.

Getting funded once is fairly common in Physical AI today. Becoming a company that investors keep financing at scale is much harder.

Who actually holds most of the money in Physical AI?

Physical AI capital is heavily concentrated around a small group of companies, with the top ten absorbing 64.5% of everything we recorded.

Waymo sits far ahead with $21.6 billion across two qualifying financings. Anduril follows at $5 billion, then Saronic Technologies at $2.525 billion, Wayve at $2.31 billion and Shield AI at $2.24 billion.

Those five companies alone account for 53.5% of the market.

The pattern is especially telling because several leaders raised repeatedly. Saronic appears three times. Wayve appears three times. Apptronik appears four times. Physical Intelligence appears three times.

Investors are concentrating money in companies that repeatedly clear the next financing hurdle, rather than spreading large checks evenly across the sector.

That helps explain why the Physical AI market can simultaneously contain nearly 300 funded companies and still feel dominated by a fairly short list of names.

Company	Qualifying rounds	Recorded capital
Waymo	2	$21.60B
Anduril	1	$5.00B
Saronic Technologies	3	$2.525B
Wayve	3	$2.310B
Shield AI	2	$2.240B
Skild AI	2	$1.700B
NEURA Robotics	2	$1.524B
Quantum Systems	2	$1.381B
Apptronik	4	$1.254B
Physical Intelligence	3	$1.070B

Is Physical AI basically a Waymo funding story?

Waymo has an enormous effect on Physical AI funding statistics, but plenty of serious capital remains once we look beyond it.

Waymo accounts for 34.3% of all recorded capital in our window. Its $16 billion financing alone represents about a quarter of the entire market.

The scale makes more sense when we look at what Waymo was already operating by the time it raised that money. In its 2026 funding announcement, the company said it had delivered 15 million rides during 2025, was providing more than 400,000 rides a week across six major U.S. metropolitan areas and planned to expand into more than 20 additional cities.

That is a very different financing problem from funding a robotics prototype. Waymo is financing vehicles, operations, safety systems, international expansion and fleet growth on top of its core AI stack.

Even after taking Waymo out, however, more than $41 billion of recorded Physical AI funding remains. Anduril, Saronic, Shield AI, Quantum Systems and the fast-growing humanoid and robot-foundation-model groups are all raising substantial amounts themselves.

Waymo therefore exaggerates the market's concentration, but it does not explain the whole thing.

What can a normal Physical AI startup realistically raise?

For a normal Physical AI startup, a round in the tens of millions is far more representative than the giant checks that dominate the headlines.

The $28 million median gives a much better sense of the market than the $179 million average. Around 40% of recorded financings are below $20 million, and another large group sits between $20 million and $50 million.

At the other end, 97 financings of $100 million or more account for $56.78 billion, or just over 90% of all recorded capital.

That is the central financial split in Physical AI these days. A founder can reasonably raise $5 million, $15 million or $30 million without being anywhere near the companies consuming most of the sector's money.

Once a company starts asking for $100 million or more, investors are usually underwriting a different problem: manufacturing, deployment, fleets, large training infrastructure, defense production or a potential general-purpose platform.

As seen above, the median remains the more useful benchmark for the ordinary company. The market average mainly tells us how extreme the top end has become.

Which part of Physical AI is getting the most funding?

Autonomous Vehicles currently dominate Physical AI funding, taking 43.2% of all recorded capital from fewer than 10% of the deals.

We count $27.18 billion across 35 Autonomous Vehicle financings. Drones & Mobile Autonomy comes next with $15.44 billion, followed by Humanoid Robotics at $8.58 billion and Robotics Foundation Models at $5.52 billion.

Industrial & Warehouse Robotics flips the pattern around. It generates more deals than any other category, with 101 financings across 85 companies, but only $4.78 billion of capital.

The difference becomes clear when we look at what these companies are trying to build.

Autonomous-driving companies have to fund AI development alongside vehicles, sensor systems, testing, safety work, fleet operations and geographic expansion. Wayve's $1.05 billion Series C, for example, was raised to push its embodied driving technology toward production vehicles and deeper work with automakers.

Industrial robotics contains far more specialized businesses that can reach real customers without first financing a global mobility network.

The money is following capital requirements as much as technological excitement.

Physical AI category	Deals	Companies	Capital	Median round
Autonomous Vehicles	35	29	$27.18B	$82.6M
Drones & Mobile Autonomy	86	74	$15.44B	$25.0M
Humanoid Robotics	57	36	$8.58B	$69.1M
Robotics Foundation Models	34	26	$5.52B	$95.5M
Industrial & Warehouse Robotics	101	85	$4.78B	$20.0M
Physical AI Control Software	39	37	$805M	$11.9M
Simulation & Synthetic Data	7	6	$641M	$89.5M

Why are autonomous-vehicle companies raising so much more than everyone else?

Autonomous-vehicle companies need much bigger checks because they are funding real transportation systems at scale, not just AI development.

Their share of Physical AI capital is 4.43 times their share of deals, by far the highest funding intensity of any major category we track.

Waymo makes the extreme case obvious. Its recent financing supports expansion into many more cities while the company is already running hundreds of thousands of weekly rides.

Wayve shows another version of the same economics. Its approach is much more model-centric, but the company still needs to work through production vehicles, automaker integrations, testing programs and deployment across different countries. After its large Series C, Wayve expanded testing into Germany and Japan and deepened relationships with global vehicle manufacturers.

Autonomous driving therefore combines the cost structure of frontier AI with the cost structure of a transportation business.

That combination explains why a relatively small number of Autonomous Vehicle deals can swallow almost half the capital in the entire Physical AI market.

Are humanoid robots actually attracting serious money now?

Humanoid Robotics has become one of the deepest funding pools in Physical AI, with $8.58 billion spread across 36 companies rather than concentrated in one obvious winner.

The category produced 57 qualifying financings, and its median round is $69.1 million.

We also do not see the kind of single-company dominance that appears in Autonomous Vehicles. The largest humanoid financing accounts for only 16.3% of the category's capital.

Recent company activity makes the numbers feel much less theoretical than they would have two years ago. Apptronik expanded its Series A beyond $935 million by early 2026, saying the money would fund Apollo production, commercial deployments, robot-training facilities and data collection. Figure's large Series C similarly identified manufacturing, GPU infrastructure and real-world data collection as core uses of capital.

Those are expensive programs, and they are happening across several companies at once.

The remarkable part is how young the sector remains. The median funded humanoid company in our research is only about two years old.

So the funding has moved very quickly. What still has to catch up is the mature financing ladder: classified Seed and Series A rounds remain far more common than Series C, D+ or Growth rounds.

Why are robotics foundation-model startups raising huge rounds so early?

Robotics Foundation Model companies are getting some of the largest early checks in Physical AI because investors are trying to finance general-purpose robot intelligence before the category has had time to mature.

The median company is only about one year old at financing, yet the median round is $95.5 million.

Skild AI makes the current logic particularly clear. Founded in 2023, the company raised $1.4 billion in early 2026 to keep building what it describes as a single general-purpose brain that can control many kinds of robots. By September, Skild was showing its S1 model learning new long-horizon tasks from a single video demonstration.

The infrastructure behind this work is also getting heavier. NVIDIA's latest Physical AI Data Factory architecture is built around mass data processing, synthetic-data generation, reinforcement learning and model evaluation. AWS and NVIDIA have meanwhile announced plans for another two million NVIDIA GPUs across AWS infrastructure, explicitly including Physical AI workloads.

A young robotics company trying to compete at that layer may need large-scale compute, simulation, data pipelines and access to many physical machines long before it looks like a normal late-stage company.

That explains the enormous early rounds. It does not tell us yet which of these companies will grow into durable businesses.

Is Physical AI still an early-stage market?

Physical AI is early-stage by deal count but already looks like a scale-up market when we follow the dollars.

Seed and Series A make up 181 of the 306 financings with a usable stage classification. Together, that is 59.2% of classified deals.

Their share of classified capital is only about 13.7%.

Series C, Series D+ and Growth Equity contain far fewer transactions but vastly more money. Series C contributes $14.77 billion, D+ another $14.18 billion and Growth Equity $16.87 billion.

Calling the whole Physical AI market either "early" or "mature" misses the split. Both descriptions can be true depending on what we are counting.

There are still many young companies entering the market. At the same time, the companies that already have credible deployment paths are consuming enormous amounts of capital.

Stage	Deals	Companies	Capital	Median round
Seed	98	95	$1.92B	$7.65M
Series A	83	74	$6.26B	$29.5M
Series B	57	55	$5.64B	$52.0M
Series C	31	31	$14.77B	$113M
Series D+	26	19	$14.18B	$121.5M
Growth Equity	11	11	$16.87B	$32.0M
Unknown	53	50	$3.31B	$15.1M

Where does Physical AI funding get hardest?

The toughest conventional step in Physical AI funding currently appears between Series B and Series C.

We count 57 Series B financings and only 31 Series C rounds. At the same time, the median check jumps from $52 million to $113 million.

That is a demanding combination. Fewer financings happen just as the amount of capital companies need roughly doubles.

By then, investors are usually looking for much more than an interesting robot or a convincing demo. Manufacturing, commercial deployments, customer economics, repeatability and the ability to operate hardware reliably start carrying much more weight.

The defense-autonomy companies show what scale can look like on the other side of that hurdle. Shield AI announced a $1.5 billion Series G in 2026 alongside $500 million of preferred equity as it expanded both its autonomy platform and broader operating footprint. Quantum Systems has used large later-stage rounds to scale production, acquire businesses and move beyond aerial systems into a wider unmanned portfolio.

For younger categories such as Humanoid Robotics and Robotics Foundation Models, that mature population barely exists yet.

The next few financing cycles should tell us much more than today's Seed and Series A activity about which Physical AI categories have real staying power.

How old are Physical AI companies when investors start writing big checks?

Physical AI companies usually need several years before they reach large-round territory, although humanoids and robot-foundation-model startups are breaking that pattern.

Across the 299 financings where we could establish company age reliably, the median company is four years old when it raises.

Companies receiving rounds above $50 million have a median age of about 5.5 years, compared with three years for those raising below $5 million.

The stage progression is fairly clean. Seed companies have a median age of two years, Series A companies three, Series B and Series C companies six, and Series D+ companies around nine.

Then we hit the new AI-native robotics groups.

Humanoid companies have a median age of only two years. Robotics Foundation Model companies come in at roughly one.

That is unusually fast capital formation for physical technology. Investors are clearly willing, for now, to fund some of these companies on the potential value of a general-purpose platform before they accumulate the operating history we would normally expect from companies raising checks of this size.

Is Physical AI funding mostly going to new companies or existing winners?

Physical AI funding is overwhelmingly flowing into follow-on rounds, which tells us that investors are spending far more on scaling selected companies than on opening the door to new ones.

We classify 291 financings as follow-ons and 66 as first financings.

The typical check also changes sharply. A first financing has a median size of $11.8 million, compared with $38 million for a follow-on.

In aggregate, follow-on rounds account for about $59.92 billion of recorded capital.

The company data tell the same story from another angle. Only a minority of companies raises more than once, yet that minority captures most of the money.

Physical AI currently has no shortage of new experiments. The much more selective part begins after the first round, when investors decide which companies deserve another $50 million, $200 million or $1 billion.

Does the United States dominate Physical AI funding?

The United States dominates Physical AI capital today much more strongly than it dominates deal creation.

North America accounts for 47.6% of financings but 75.9% of recorded capital. The United States alone contributes about $46.71 billion across 162 deals.

Asia-Pacific is much broader than its dollar share suggests. It produces 28.7% of deals but 11.1% of capital. Europe contributes 21.7% of deals and 12.8% of capital.

Part of the American lead comes from the obvious giants: Waymo, Anduril, Shield AI, Skild AI, Apptronik and several other heavily funded companies are based there.

Yet the difference remains visible in ordinary round sizes too. North America's median financing is $34 million. Asia-Pacific is close behind at $28.9 million, while Europe's is $13.9 million.

Europe does have notable exceptions. Germany's NEURA Robotics and Quantum Systems have both raised large amounts, and the United Kingdom's Wayve is one of the world's best-funded autonomous-driving AI companies.

Still, the deepest pool of large-scale private Physical AI capital currently sits in the United States.

Which investors keep coming back to Physical AI?

Khosla Ventures shows up more often as a lead or co-lead than any other financial investor in our Physical AI research, but the market is much less clubby than the giant rounds might suggest.

Khosla appears seven times across Autonomous Vehicles, Robotics Foundation Models, industrial robotics, drones and humanoids. Lightspeed appears five times, while Eclipse, IDG Capital and Plural each appear four times.

After that, the recurring list broadens quickly. Sequoia Capital, Andreessen Horowitz, Thrive Capital, SoftBank and Kleiner Perkins each appear three times.

Across the market, we identify 330 distinct disclosed lead or co-lead investors. Only 40 repeat.

That fragmentation tells us something useful. Physical AI has attracted mainstream venture firms, deep-tech specialists, growth investors, sovereign money and industrial capital simultaneously. No tiny group of VCs is underwriting the whole category.

We should also be careful with investor-related dollar totals. If a firm led a $1 billion round, that does not mean the firm itself invested $1 billion. Round size measures the financing it helped lead, not its individual check.

Frequency and breadth are more useful for seeing who really keeps showing up.

Do corporate investors matter more in Physical AI than in normal software?

Corporate and strategic investors matter a lot in Physical AI because companies often need infrastructure, factories, vehicles, deployment sites and real-world data alongside money.

We identify strategic participation in 44% of all financings.

The difference in round size is striking. Financings with a strategic investor have a median size of $69.1 million, compared with $15.1 million when no strategic investor is identified.

Selection clearly plays a role here: companies raising bigger rounds are more likely to attract corporate partners anyway. But the relationships themselves are becoming much deeper than a logo on a cap table.

AWS and NEURA Robotics, for example, are working together on robot training, fleet intelligence and potential fulfillment-center deployments. NVIDIA now works across a large robotics network that includes ABB, FANUC, KUKA, Figure, Skild AI and others, providing everything from simulation to models and edge computing.

Apptronik combines venture backing with Google, Mercedes-Benz and John Deere. Wayve has brought in NVIDIA, Microsoft, SoftBank and Uber. Shield AI's financing has included strategic defense and industrial investors.

Physical AI companies need places to train machines, manufacture them and prove that they work. A strategic backer can sometimes provide all three.

That makes corporate participation particularly useful in this market once companies move from research into deployment.

Is Physical AI funding still accelerating, or are giant rounds making it look hotter than it is?

Physical AI deal activity remains healthy today, but the spectacular monthly funding totals mostly come from huge individual financings rather than an equally spectacular increase in the number of deals.

Across the 33 completed months in our research window, the median month contains ten financings. The average is 10.9.

The busiest months reached 23 deals. That is strong activity, but only a little more than twice the normal monthly pace.

Capital behaves very differently. The median completed month contains about $654 million of recorded funding, while the biggest month reached $18.45 billion.

Waymo's $16 billion round created most of that spike by itself.

A different kind of boom appeared in April 2026, when we recorded 23 deals but no single transaction above $200 million. That was genuine breadth: lots of companies were getting funded at once.

The distinction is worth keeping. Physical AI is clearly an active funding market these days, and the newest operating announcements from Skild AI, NVIDIA, NEURA, Apptronik and others show that capital is being put to work. But a monthly dollar chart can still make the market look far more explosive than the underlying deal flow really is.

What is Physical AI funding still missing?

Physical AI funding still lacks a deep late-stage population across its newest categories, and that is probably the biggest weakness hidden beneath today's huge totals.

Robotics Foundation Models have 34 financings and billions of dollars of capital, yet their classified ladder contains almost no traditional later-stage depth. Humanoids show a similar pattern: lots of Seed and Series A activity, followed by a much smaller population at C and beyond.

Some of that is simply a matter of time. Companies founded in 2023 cannot magically produce a ten-year financing history by 2026.

There are other gaps too. The Middle East produces only six deals in our dataset and none above $50 million. Physical AI Control Software has 39 financings across 37 companies but only $805 million of total recorded capital and no classified Series D+ round.

Repeat financing is also rare for most companies. More than four-fifths of funded companies appear only once.

These gaps show where the hype has moved faster than the financing system underneath it.

The next test is whether today's huge early checks produce a much deeper population of Series C, D and Growth companies, especially in humanoids and general-purpose robot intelligence.

What should we actually take away from Physical AI funding today?

Physical AI is one of the few technology markets where a genuinely broad startup ecosystem and extreme capital concentration coexist at the same time.

Nearly 300 companies received qualifying funding in our research window, so this is much larger than a handful of famous robotics names. Serious money, however, accumulates quickly around a much smaller group of repeat winners.

The market also has two very different financing speeds. Specialized robotics and early autonomy companies often raise conventional venture rounds in the low tens of millions. Autonomous driving, defense autonomy, humanoids and robot foundation models can move into nine-figure or billion-dollar territory surprisingly quickly.

Company age tells us how unusual the current cycle is. Most Physical AI companies still get larger checks as they mature, but humanoids and robot-foundation-model companies are attracting major capital after only one or two years.

The hardest part of the market starts after the early rounds. Series B to Series C is where the financing population contracts noticeably while the typical check jumps above $100 million. Plenty of companies can raise enough to build something; far fewer receive enough capital to industrialize it.

Strategic capital is another defining feature. Compute providers, manufacturers, automakers, logistics operators and defense groups keep appearing because deploying Physical AI requires physical infrastructure and customers as much as financial backing.

For now, the clearest picture is a broad experimental base feeding a much narrower industrial-scale layer. The companies that reach that upper layer gain access to an extraordinary amount of capital. The real question for the next few years is how many of today's heavily funded young companies can actually stay there.