What are the top robotics startups by revenue today?

The ranking of top startups in the robotics market by revenue

Below is a table ranking all the companies in this market by their current revenue scale. You can find our methodology at the end of this page.

If you want a deeper understanding of the market and its current dynamics, get our report covering the Robotics Market.

What actually counts as a robotics startup today?

For this ranking, we count private, venture-backed companies whose core business is selling robots, robotic systems, or software that directly operates robots.

That definition removes several businesses that often appear in robotics rankings. FANUC, ABB Robotics, Yaskawa and KUKA are mature industrial groups. Symbotic and PROCEPT BioRobotics are public. Boston Dynamics is controlled by Hyundai. Geek+, DOBOT and Unitree have also moved beyond the private-startup category.

Unitree is a good example of why the distinction matters. Shanghai Stock Exchange material shows that the company generated RMB1.70 billion of revenue in 2025, up from RMB393 million in 2024. Humanoid robots alone contributed RMB868 million. Those numbers would put Unitree very high in a broader robotics-company ranking, but its public listing means we no longer treat it as a startup here.

We use a fairly broad definition of robotics, so autonomous aircraft and vessels can qualify when the physical system itself is central to the business. That brings companies such as Shield AI and Saronic into the research, although their revenue is less directly comparable with warehouse or industrial robotics.

Which robotics startups make the most revenue right now?

Exotec, Agile Robots and Hai Robotics form the clearest top group among private robotics startups with strong current revenue evidence.

French filed financial statements put Exotec at roughly €343 million of 2025 revenue. Agile Robots generated €300 million according to CEO Zhaopeng Chen in a recent Wall Street Journal interview. Hai Robotics reported RMB2.02 billion of 2025 revenue in its Hong Kong listing application.

A few companies may be larger depending on which estimates we accept. Shield AI has been estimated around $400 million of annual revenue. Saronic has been estimated around $200 million. ANYbotics also has outside estimates placing it well into nine figures. Those figures are useful, but they should not get the same weight as filed accounts or regulatory disclosures.

Locus Robotics is another important case because its roughly $180 million figure is ARR rather than annual recognized revenue. Skild AI has now crossed $100 million of ARR, while Carbon Robotics says it generates more than $100 million in annual revenue.

Rank	Robotics startup	Best current commercial figure	Metric	How solid is the figure?
1	Exotec	~€343M	2025 annual revenue	Filed financial statements
2	Agile Robots	€300M	2025 annual revenue	CEO disclosure to WSJ
3	Hai Robotics	RMB2.02B	2025 annual revenue	HKEX filing
4	Shield AI	~$400M	Estimated annual revenue	Third-party estimate
5	Saronic	~$200M	Estimated annual revenue	Third-party estimate
6	Locus Robotics	~$180M	ARR	Specialist estimate
7	ANYbotics	~$177M	Estimated annual revenue	Third-party estimate
8	AgiBot	RMB1.05B	2025 annual revenue	Chairman disclosure
9	Keenon Robotics	~$139M	Estimated annual revenue	Third-party estimate
10	Carbon Robotics	>$100M	Annual revenue	Company disclosure
11	Skild AI	>$100M	ARR	Company disclosure

Is Exotec really the biggest private robotics startup by revenue?

Exotec has the strongest case for number one today because its roughly €343 million revenue figure combines large scale with unusually solid financial evidence.

The company had already crossed $1 billion of cumulative sales by 2024 and says more than 10,000 robots have now been manufactured and deployed across more than 200 customer sites. Those operating numbers fit the scale visible in its filed accounts.

Exotec also keeps adding large deployments. A recent installation with GXO for Guess in the Netherlands uses 127 robots and can process up to 70,000 pieces per day. Earlier deployments span customers including Carrefour, Decathlon, Gap, Renault and Uniqlo.

Shield AI's estimated revenue could be higher, but we do not have an equally strong public financial disclosure. Agile Robots is close, but its €300 million figure comes from its CEO rather than filed consolidated accounts.

For now, Exotec has the cleanest combination of current revenue, source quality and robotics-specific business scope at the top of the ranking.

How big is Agile Robots now?

Agile Robots is already a €300 million business, and its CEO says the company expects revenue to roughly double this year.

The €300 million figure came directly from CEO Zhaopeng Chen in The Wall Street Journal. Much of this scale has come from industrial automation rather than humanoid robots.

Agile has bought more than a dozen companies to build manufacturing, integration and factory-floor expertise. According to Chen, acquired businesses now contribute roughly 10% to 15% of new contracts. Its humanoid robots are still being tested in pilots, while the established industrial side is already bringing in hundreds of millions of euros.

That gives Agile a very different financial profile from the most heavily funded humanoid startups. It already has a sizeable operating business and is using that industrial base to move into more flexible AI-powered robotics.

The company's stated plan to double revenue is worth watching, but we keep the ranking on the €300 million already generated rather than a future target.

Is Hai Robotics already bigger than most U.S. robotics startups?

Yes. Hai Robotics is already larger by reported revenue than almost every private U.S. robotics startup for which we have a comparable figure.

Its Hong Kong listing application shows revenue rising from RMB807 million in 2023 to RMB1.36 billion in 2024 and RMB2.02 billion in 2025. That is roughly 150% growth in two years.

The company then generated RMB1.12 billion in the first half of the following year, so the business kept growing rather than posting one unusually strong annual number.

Hai is also becoming much more international. Revenue from outside mainland China rose from about one-quarter of the business in 2023 to roughly half in the latest half-year period. More than half of new orders were coming from overseas customers.

Hai is no longer just a domestic Chinese warehouse-robot vendor. It is turning into one of the largest global private robotics companies.

Does Shield AI really belong near the top of a robotics revenue ranking?

Shield AI belongs near the top under a broad robotics definition, but its roughly $400 million revenue estimate comes with much more uncertainty than the figures above it.

Shield AI builds autonomous aircraft and the Hivemind autonomy platform. Some of that business looks like classic robotics hardware, while some looks more like autonomy software and defense systems.

That broader mix makes the revenue perimeter harder to compare with Exotec, Hai Robotics or Carbon Robotics, where the link to physical robotic systems is much clearer.

The second issue is source quality. The roughly $400 million figure circulating around Shield AI is an estimate rather than a filed company disclosure.

Shield AI clearly belongs in the group of commercially large private physical-autonomy companies. We just would not treat $400 million as precisely comparable with €343 million of filed Exotec revenue.

How much revenue is Saronic making?

Saronic appears to have reached roughly $200 million of annual revenue, although the current figure is still based on outside estimates.

That would be an enormous jump from the much smaller scale reported a year earlier and fits the rapid growth in autonomous-vessel contracts around the company.

Saronic has secured hundreds of millions of dollars of production work and raised enough capital to build far more manufacturing capacity. Those contracts make a nine-figure revenue business plausible.

We still separate contracts from recognized revenue. Defense robotics companies can announce very large awards that are delivered over several years, and the headline contract value can easily make the company look larger than its current income statement.

The best evidence today puts Saronic in the second group behind the clearest $250 million-plus private companies.

Is Locus Robotics bigger than most humanoid robot startups?

Yes. Locus Robotics is already commercially bigger than almost every famous private humanoid company we can measure.

Specialist estimates put Locus around $180 million of ARR. The company crossed $100 million ARR years ago and has kept expanding its Robots-as-a-Service business across fulfillment centers.

That subscription model gives Locus a mature revenue profile for robotics. Customers keep paying for deployed fleets rather than generating revenue only when a new robot is sold.

Compare that with Agility Robotics. Recent SEC material shows Agility generated about $1.8 million of revenue in 2025 even though the company has now secured more than $300 million of multi-year Digit orders.

Figure AI and Apptronik have major factory partnerships and huge valuations but still do not publish comparable revenue figures.

Warehouse robotics gets much less attention than humanoids these days. The commercial gap is still huge.

Is ANYbotics really making more than $100 million a year?

ANYbotics is probably a large industrial-robotics business now, but its exact revenue remains one of the shakier numbers in the upper ranking.

One commercial database estimates around $177 million of 2025 revenue, while other sources have placed the company considerably lower. That spread is too wide for us to treat $177 million as established fact.

The operating evidence is easier to trust. ANYmal robots are deployed for industrial inspection in oil and gas, chemicals, utilities, mining and other asset-heavy industries. ANYbotics has also been building a larger global sales organization as it moves beyond early deployments.

So we are comfortable saying ANYbotics has reached meaningful commercial scale. Its exact position among the nine-figure robotics companies is still much less certain than Exotec, Agile Robots or Hai Robotics.

How fast is AgiBot turning humanoid robots into revenue?

AgiBot has one of the fastest revenue ramps anywhere in humanoid robotics, with reported sales jumping from RMB60 million to RMB1.05 billion in one year.

Chairman Deng Taihua disclosed the RMB1.05 billion figure for 2025. That works out to roughly 17 times the previous year's revenue.

AgiBot has also moved into much higher production volumes. The company has reported producing thousands of embodied robots across several product lines rather than keeping commercialization limited to a handful of factory pilots.

This is one place where the Chinese humanoid market looks very different from the U.S. market. AgiBot already has meaningful reported sales, while many high-profile American humanoid companies are still measured mostly through funding, pilots and future orders.

The company's next revenue targets are much larger, but the RMB1.05 billion already achieved is enough to put AgiBot among today's largest private robotics startups.

How much revenue does Keenon Robotics make?

Keenon Robotics appears to be a $100 million-plus service-robot company, although its exact revenue is still estimated rather than formally disclosed.

One widely used company database puts 2025 revenue around $139 million. Keenon's long operating history makes that scale believable: its robots have been deployed for years in restaurants, hotels, hospitals and other indoor service environments.

The company also benefits from a market that reached commercialization earlier than humanoids. Delivery and hospitality robots have narrow jobs, predictable environments and relatively obvious customer economics.

We therefore keep Keenon in the upper part of the ranking, but with lower confidence than AgiBot or Carbon Robotics. A proper company filing could easily shift its position by several places.

Has Carbon Robotics really crossed $100 million in annual revenue?

Yes. Carbon Robotics says it now generates more than $100 million in annual revenue.

The company builds the LaserWeeder and tractor-autonomy products for agriculture and operates across 15 countries. Its revenue figure comes directly from Carbon rather than from an external estimate.

Carbon is a useful example of what is working commercially in robotics right now. Laser weeding tackles a specific farm expense, so growers can compare the cost of the machine with labor, herbicide use and crop economics.

That makes buying decisions much easier than for a general-purpose robot with an uncertain set of future jobs.

Carbon has quietly become one of the stronger private robotics businesses in the U.S., even though it receives much less attention than humanoid startups valued at several billion dollars.

Is Skild AI's $100 million ARR as impressive as it sounds?

Yes. Skild AI crossing $100 million ARR is one of the strongest new commercialization signals in robotics, especially because the company also disclosed $50 million of recognized revenue since deployments began.

Skild says it passed $100 million ARR only ten months after its first commercial deployment and now has more than 60 paying customers.

The revenue mix is revealing. Roughly 90% comes from manipulation work, while mobility accounts for only about 10%. Customers are using Skild's software for tasks involving warehouses, factories, food preparation, security, inspection and data centers.

ARR is still different from annual recognized revenue, so we do not rank Skild as if it had already completed a $100 million revenue year.

Even with that distinction, very few robotics companies have gone from first commercial deployment to this level of recurring business so quickly.

Could Pudu Robotics actually belong near the top of this ranking?

Probably, but Pudu Robotics still does not disclose enough revenue detail for a defensible exact position.

Pudu says its revenue more than doubled in 2025 and cumulative shipments passed 120,000 robots. More recently, the company said it had over 130,000 robots operating across more than 80 countries and regions.

Its industrial-delivery business has also kept accelerating. Pudu reported that first-quarter industrial robot revenue doubled year on year, while the T300 AMR passed 4,000 units shipped within its first year and the newer T150 exceeded 1,000 units within four months.

Those numbers clearly describe a large commercial robotics business.

What they do not give us is annual company revenue. Multiplying shipment numbers by guessed selling prices would produce a neat-looking number with very weak foundations, so we leave Pudu unranked rather than manufacture precision.

Where are Figure AI and Apptronik in the robotics revenue ranking?

Figure AI and Apptronik remain impossible to rank confidently by revenue because neither company publicly discloses a usable company-wide revenue figure.

Figure has real operating evidence. BMW says Figure 02 worked for about ten months at its Spartanburg plant, supported production of more than 30,000 BMW X3 vehicles, moved more than 90,000 components and accumulated roughly 1,250 operating hours.

Figure is now moving to a newer robot generation with BMW, so this has progressed beyond a one-day demonstration.

Apptronik has also attracted serious industrial partners. Its expanded Series A reached $935 million, and reporting around the round put the company's valuation at roughly $5.3 billion.

Those facts tell us both companies have strong investors, real customers and credible deployment programs. They still do not tell us how much revenue the companies make.

That is why neither sits near the top of a revenue ranking today, even though both rank near the top by attention and valuation.

Is Agility Robotics' $300 million order book already real revenue?

No. Agility Robotics has more than $300 million of multi-year Digit orders, while SEC material shows only about $1.8 million of 2025 revenue.

That gap is one of the clearest examples of why robotics rankings can become misleading.

Agility's order book covers roughly 1,000 Digit v5 robots under multi-year arrangements, largely using a Robots-as-a-Service model. The company itself tells investors that these orders represent potential value to be recognized over time rather than current-period revenue.

The commercial progress is still substantial. Digit has been deployed with customers including GXO, Schaeffler, Toyota and Amazon, and Agility says earlier Digit systems accumulated more than 65,000 hours of operation.

But $300 million in contracted future value and $1.8 million of recognized annual sales describe very different stages of commercialization.

For a revenue ranking, the $1.8 million figure is the one that matters today.

Why are warehouse robotics startups making so much more money than humanoid startups?

Warehouse robotics companies started solving expensive, repetitive customer problems years earlier, so they have had much more time to build repeatable revenue.

Exotec sells systems around storage density, picking and throughput. Hai Robotics automates case handling. Locus sells mobile fulfillment capacity as a recurring service.

Customers can measure those products against warehouse labor, picking speed, order accuracy and space utilization. The buying decision is fairly concrete.

Humanoid robots have a wider theoretical market, but most customers are still figuring out which tasks justify large-scale deployment. Even impressive factory pilots can take years to turn into thousands of paid robots.

This is why a warehouse company such as Exotec can generate hundreds of millions of euros while companies commanding far higher humanoid valuations may disclose little revenue at all.

Company	Main commercial model	Current scale evidence	Typical customer problem
Exotec	Integrated warehouse systems	~€343M annual revenue	Storage and fulfillment
Hai Robotics	Warehouse robot systems	RMB2.02B annual revenue	Case handling
Locus Robotics	Robots-as-a-Service	~$180M ARR	Fulfillment labor
Carbon Robotics	Equipment sales	>$100M annual revenue	Weed control
Skild AI	Robotics software	>$100M ARR	Manipulation and robot intelligence
Agility Robotics	Humanoid RaaS / sales	~$1.8M 2025 revenue, $300M+ orders	Industrial material handling
Figure AI	Humanoid deployment	Revenue undisclosed	Factory manipulation

Are humanoid robotics startups finally catching warehouse robotics on revenue?

Some Chinese humanoid companies are getting close, but humanoid robotics still trails warehouse automation on proven private-company revenue overall.

AgiBot already reported RMB1.05 billion of annual revenue. Before going public, Unitree generated RMB1.70 billion of total revenue, with RMB868 million coming from humanoid robots alone.

Those are serious numbers.

The U.S. picture is much earlier. Agility's recognized revenue was only about $1.8 million in 2025 despite its large order book. Figure and Apptronik do not disclose comparable revenue. Their commercialization is visible through factory deployments rather than income statements.

China therefore has the clearest examples of humanoid companies already translating unit shipments into substantial revenue.

Warehouse robotics still has more companies operating at large, proven commercial scale, but the gap is beginning to narrow.

Are Chinese robotics companies more commercial than U.S. robotics startups now?

China has a deeper pool of robotics companies already shipping large volumes, while the U.S. has more private companies attracting huge amounts of capital before reaching the same revenue scale.

Hai Robotics generated RMB2.02 billion in 2025. AgiBot reached RMB1.05 billion. Pudu has shipped more than 130,000 robots. Unitree generated RMB1.70 billion before becoming public.

Several other large Chinese robotics businesses have also already listed, including Geek+ and DOBOT. Geek+ reported RMB3.17 billion of 2025 revenue, up 31.6%, and moved into adjusted profitability.

The U.S. still has plenty of real commercial scale through companies such as Locus, Carbon Robotics, Shield AI and Saronic. Yet the highest-profile American humanoid companies are generally earlier in revenue conversion than their valuations suggest.

Europe is easy to overlook here. Exotec and Agile Robots alone give the region two of the largest private robotics companies by current revenue.

There is no single geographic winner. China stands out for manufacturing volume and the depth of commercially scaled robot makers; Europe has unusually strong private industrial leaders; the U.S. has the biggest cluster of richly funded bets on the next generation of robotics.

Which robotics startups have revenue numbers we can actually trust?

Exotec, Hai Robotics, Carbon Robotics, AgiBot and Skild AI currently have some of the most useful private-company commercial disclosures in the market.

Hai's figures come through regulatory filing material. Exotec has filed financial accounts. Carbon directly states more than $100 million of annual revenue. AgiBot's chairman has given a specific annual figure. Skild published both ARR and recognized revenue rather than giving only one flattering run-rate metric.

Agile Robots also gives us a clear annual figure, although it comes from a CEO interview rather than filed consolidated statements.

The weaker part of the ranking starts when we reach figures for Shield AI, Saronic, ANYbotics and Keenon. These companies may be exactly as large as the estimates suggest, but the underlying evidence is thinner.

That difference in confidence is important enough to show directly.

Company	Figure used	Metric	Confidence in current scale
Exotec	~€343M	Annual revenue	High
Hai Robotics	RMB2.02B	Annual revenue	High
Agile Robots	€300M	Annual revenue	High
AgiBot	RMB1.05B	Annual revenue	High
Carbon Robotics	>$100M	Annual revenue	High
Skild AI	>$100M	ARR	High on ARR, less comparable with annual revenue
Locus Robotics	~$180M	ARR estimate	Medium
Saronic	~$200M	Revenue estimate	Low
ANYbotics	~$177M	Revenue estimate	Low
Keenon Robotics	~$139M	Revenue estimate	Low
Shield AI	~$400M	Revenue estimate	Low

Does going public completely change the robotics startup leaderboard?

Yes. Once recently listed robotics companies are removed, the private-company ranking looks dramatically different.

Geek+ reported RMB3.17 billion of 2025 revenue. Unitree reported RMB1.70 billion. DOBOT reported about RMB492 million. All three would sit high in the table if we ignored listing status.

Geek+ alone would be larger than Exotec on revenue.

This explains why older "top robotics startup" lists can age badly. A company can still feel young and venture-backed long after it has crossed into the public markets.

Company	Latest major revenue figure	Status now	Included here?
Geek+	RMB3.17B	Public	No
Unitree Robotics	RMB1.70B	Public	No
DOBOT	~RMB492M	Public	No
Exotec	~€343M	Private	Yes
Agile Robots	€300M	Private	Yes
Hai Robotics	RMB2.02B	Private / IPO applicant	Yes
AgiBot	RMB1.05B	Private	Yes
Carbon Robotics	>$100M	Private	Yes

Which kinds of robotics startups are actually making the most money?

Warehouse and industrial robotics still generate the deepest pool of proven private-company revenue.

Exotec, Hai Robotics and Locus together represent several hundred million dollars of annual revenue or ARR. Adding Agile Robots pushes the industrial and warehouse cluster comfortably above $1 billion in current commercial scale.

Defense autonomy is growing quickly through Shield AI and Saronic, although their exact revenue is less transparent.

Agricultural robotics now has a clear nine-figure company in Carbon Robotics.

Humanoids are improving faster than any other category, especially in China. AgiBot has already crossed RMB1 billion, while public companies Unitree and UBTECH show that substantial humanoid revenue is possible.

Still, most of the money today comes from robots built around specific jobs: warehouse picking, factory automation, crop treatment, industrial inspection and autonomous defense systems.

General-purpose robots may eventually become much larger. The current revenue table says specialization is still paying the bills.

Are robotics startups actually becoming real businesses now?

Yes. The market now has a meaningful group of private robotics companies generating nine-figure annual revenue or ARR.

Exotec is above €300 million. Agile Robots is at €300 million. Hai Robotics is above RMB2 billion. Locus is estimated around $180 million ARR. AgiBot reported RMB1.05 billion. Carbon has passed $100 million in annual revenue. Skild crossed $100 million ARR within ten months of launching commercially.

Those companies also use very different business models. Some sell large hardware systems. Others charge recurring subscriptions. Defense companies work through contracts. Carbon sells capital equipment to farms. Skild sells robot intelligence.

So robotics is no longer one big pre-revenue technology bet.

There are still plenty of spectacular demos, very high valuations and companies with tiny current sales. Alongside them, a second group has quietly built substantial businesses.

So what are the top robotics startups by revenue today?

Exotec has the strongest current claim to number one among private robotics-focused startups, with roughly €343 million of 2025 revenue, while Agile Robots and Hai Robotics make up the rest of the clearest top tier.

Shield AI could be larger at an estimated $400 million, but that figure is less reliable and covers a broader autonomy business. Saronic appears to be around $200 million. Locus is around $180 million ARR. ANYbotics may also be in that range, though estimates vary sharply.

AgiBot has become one of the biggest private humanoid companies with RMB1.05 billion of revenue. Keenon is likely above $100 million. Carbon Robotics has directly confirmed more than $100 million in annual revenue, and Skild AI recently crossed $100 million ARR.

The ranking also exposes a big disconnect in robotics right now. Figure AI and Apptronik are among the sector's highest-valued startups but still do not publish comparable revenue. Agility has more than $300 million in future orders while its latest disclosed annual revenue was only about $1.8 million.

Meanwhile, less glamorous warehouse, industrial and agricultural robotics companies are already generating hundreds of millions of dollars.

The clearest answer is that the robotics companies making the most money today are mostly the ones that have spent years solving narrow, expensive problems at scale. The humanoid race is becoming commercially real, particularly in China, but it has not yet replaced warehouse and industrial automation at the top of the private revenue leaderboard.

[PLACEHOLDER ENCART]