Which humanoid robotics company has the biggest order book?
In our humanoid robotics market deck, you will find everything you need to understand the market
SUMMARY
Agility Robotics has the biggest clearly disclosed humanoid-specific order book we can verify, with more than $300 million of contracted multi-year Digit v5 orders.
NEURA Robotics has a much larger headline number at more than $1 billion, but that figure combines its existing order book with a strategic deployment pipeline across a broader robotics portfolio. It is not a clean humanoid backlog number.
UBTECH is the closest large competitor with a comparable humanoid-specific demand figure: RMB1.4 billion, roughly $200 million, of humanoid orders reported for 2025. Its advantage is that a large part of that demand is already showing up as shipments and revenue.
The quality of an “order” matters almost as much as its size. A signed multi-year RaaS contract, a procurement order, a refundable consumer reservation and a production target can all produce big-looking numbers while representing very different levels of customer commitment.
Agility's lead is unusually easy to document because SEC materials connect the $300 million-plus figure to 1,000 Digit v5 robots under a three-year service agreement. The weak point is concentration: most of the visible order book comes from one huge customer contract.
UBTECH looks more commercially mature even though its disclosed annual order total is smaller. It sold 1,079 full-size humanoids in 2025 and another 921 in the first half of 2026, while full-size humanoid revenue reached RMB590.3 million in that six-month period.
The shipment leaders tell a different story again. AgiBot and Unitree shipped far more humanoids than Agility, but neither publishes enough current backlog detail for us to rank its unfilled orders confidently.
Figure's claimed path toward 100,000 robots and 1X's 10,000 NEO reservations are real demand clues, but neither belongs beside contracted backlog without qualification. One is prospective customer volume; the other is based on fully refundable deposits.
Tesla Optimus is even harder to compare because internal Tesla deployments are not third-party customer orders. Huge production ambitions can validate supply plans without creating an external order book at all.
The broader commercial picture therefore splits into several leaders: Agility in disclosed humanoid contract value, UBTECH in repeated conversion of orders into revenue, AgiBot in shipment volume, and NEURA in the largest overall commercial demand pool across a mixed robotics portfolio. The headline winner is Agility, but the company with the biggest visible backlog is not automatically the company furthest along commercially.
This market map, featured in our humanoid robotics market deck, highlights top companies and startups in the humanoid robotics market
Which humanoid robotics company has the biggest order book?
Agility Robotics currently has the biggest clearly disclosed humanoid-specific order book we can verify, with more than $300 million of multi-year orders for its Digit v5 humanoid.
That answer needs one important boundary. NEURA Robotics reports more than $1 billion of existing orders and strategic deployment pipeline, which is much larger. But NEURA sells several kinds of robots, from collaborative arms to autonomous mobile robots and humanoids, and it does not disclose how much of that $1 billion belongs specifically to humanoids.
Agility gives us a cleaner number. In its latest public-company filings and merger materials, the company says it has secured more than $300 million of multi-year Digit v5 orders. SEC materials give us even more detail: the headline contract involves 1,000 Digit v5 robots under a three-year Robotics-as-a-Service agreement.
UBTECH is the closest major competitor with a comparable humanoid-specific number. The company said its humanoid robot orders reached RMB1.4 billion, roughly $200 million, during 2025. Its latest financial results also show that those orders are increasingly becoming actual revenue.
So the ranking is fairly simple for the metric in the title: Agility first for disclosed humanoid-specific contracted value, NEURA first if we include the wider robotics businesses of humanoid companies, and UBTECH strongest among the companies already turning large humanoid orders into significant reported sales.
If you want more recent data on this point, please see our latest humanoid robotics market report.
Why is comparing humanoid robot order books so messy?
Humanoid robot companies use the word “order” for very different things, which makes a seemingly simple ranking surprisingly unreliable.
Agility reports contracted multi-year orders. UBTECH has reported procurement contracts and annual order totals. 1X talks about consumer reservations. Figure has discussed a path toward very large customer volumes without publishing a firm backlog. NEURA combines an order book with what it calls a strategic deployment pipeline.
Those categories can differ enormously economically. A three-year customer contract worth hundreds of millions of dollars creates a much stronger commitment than a refundable $200 deposit. A framework agreement allowing a customer to buy thousands of robots may eventually become huge while committing very little today.
Production targets create another source of confusion. Tesla can plan to manufacture enormous numbers of Optimus robots for its own factories without creating an external customer order book. Likewise, a factory capable of producing 10,000 robots a year tells us about supply capacity, not how many customers have actually agreed to buy them.
For this comparison, we count signed or explicitly reported humanoid orders most heavily. Reservations, prospective pipeline and possible future volumes are useful evidence of demand, but we keep them separate.
| Type of demand | How much should we trust it as an order book? |
|---|---|
| Signed purchase or RaaS contract | Very strong |
| Reported procurement order | Strong |
| Conditional framework agreement | Useful, but needs qualification |
| Refundable consumer reservation | Early demand evidence |
| Sales pipeline | Too preliminary to count as backlog |
| Production target | Does not measure customer demand |
As this chart shows, and as featured in our humanoid robotics market deck, search interest in where to buy robots has been rising steadily
Does NEURA Robotics really have more than $1 billion of humanoid orders?
NEURA Robotics has the biggest headline number in the sector, but we cannot support the claim that more than $1 billion represents humanoid orders.
NEURA said during its latest major financing that its “existing orderbook and strategic deployment pipeline” exceeded $1 billion. The company also raised up to $1.4 billion in its Series C, giving it one of the largest pools of capital in physical AI.
The problem is product mix. NEURA sells far more than its 4NE-1 humanoid. Its portfolio includes MAiRA collaborative robots, autonomous mobile systems and other industrial robots, while its acquisition of ek robotics expanded that non-humanoid business further.
The history of NEURA's backlog makes this especially clear. Back in 2023, NEURA was already reporting around $450 million of orders even though its commercial business was largely built around other cognitive robots rather than volume shipments of 4NE-1. By early 2025, the company said its order book had reached €1 billion.
The wording has broadened lately too. NEURA now combines its “orderbook” and “strategic deployment pipeline” in the $1 billion-plus figure. Pipeline normally includes opportunities that are less firm than booked contracts.
We therefore treat NEURA as the company with the largest overall disclosed commercial demand pool among humanoid developers. Without a product-level breakdown, putting $1 billion beside Agility's Digit backlog would give readers the wrong impression.
How solid is Agility Robotics' $300 million Digit order book?
Agility Robotics' $300 million-plus Digit order book is unusually well documented for this industry, although most of the value comes from one very large customer agreement.
Recent Agility materials describe more than $300 million of multi-year contracted orders for Digit v5. SEC disclosures go further and link the amount to 1,000 robots under a three-year Robotics-as-a-Service contract.
That works out to more than $300,000 of contracted value per robot on average. We should not interpret that as the purchase price of a Digit. The contract covers a multi-year service model, so software, support, deployment and recurring services are included in the economics.
There are conditions attached. Agility says the orders remain subject to contractual milestones, and the agreement includes customer warrants that vest as robots are deployed. So the entire $300 million should not be treated as cash already locked in regardless of what happens.
Customer concentration is the bigger weakness. One agreement accounts for most of the publicly visible order book. Agility says it also has more than 30 customers in its pipeline, but we cannot count that pipeline until those discussions turn into orders.
That still leaves Agility as the current leader in disclosed humanoid contract value. The number is solid; the diversity behind it is much less impressive.
If you want more recent data on this point, please see our latest humanoid robotics market report.
This chart, featured in our humanoid robotics market deck, illustrates yearly venture capital funding for humanoid robotics startups
Are Agility's Digit robots already doing real work for customers?
Agility Robotics has already moved Digit into real warehouses and factories, which makes its large future order more credible than an order attached only to prototypes.
Digit has worked with GXO Logistics, Schaeffler, Toyota Motor Manufacturing Canada, Mercado Libre and other large companies. Agility recently said its customer deployments across nine facilities had accumulated more than 65,000 operating hours.
GXO is particularly useful evidence because the relationship went beyond an initial pilot. After testing Digit in warehouse workflows, GXO entered a multi-year Robotics-as-a-Service agreement with Agility for commercial operations.
The work itself is fairly narrow today. Digit has focused on repetitive material-handling jobs such as moving totes between shelves, carts and conveyors. That sounds less spectacular than a general-purpose humanoid doing dozens of jobs, but repetitive material movement is exactly the kind of task companies can measure and potentially justify economically.
Agility's historical revenue remains tiny relative to the contract it is preparing to fulfill. Its public filings showed only about $1.8 million of revenue in 2025. The next generation of Digit therefore has to turn the company from a low-volume deployment business into something hundreds of times larger in commercial value.
That gap is one of the biggest execution tests anywhere in humanoid robotics right now.
Is UBTECH actually closer to mass commercialization than Agility Robotics?
UBTECH currently looks further along in converting humanoid demand into substantial revenue and physical deliveries, even though its last disclosed annual order total is smaller than Agility's headline backlog.
UBTECH said it received RMB1.4 billion, roughly $200 million, of humanoid orders during 2025. More importantly, its financial statements now show a large portion of humanoid business flowing through the income statement.
Full-size humanoid products and services generated about RMB820 million of revenue in 2025, compared with only RMB35.6 million in 2024. The company sold 1,079 full-size humanoids during the year.
The latest numbers show that growth continuing rather than disappearing after one big delivery cycle. UBTECH reported RMB590.3 million of full-size humanoid revenue in the first half of 2026, up 1,445% year over year. Sales reached 921 units in those six months.
That means UBTECH sold almost as many full-size humanoids in six months as it sold throughout the previous year. Humanoids also generated 46.5% of group revenue, making them UBTECH's largest business line.
Agility still has the cleaner claim to the biggest currently disclosed humanoid-specific contracted value. But if we ask which company has already shown that customers will repeatedly pay meaningful amounts for full-size humanoids, UBTECH has stronger financial evidence.
If you want more recent data on this point, please see our latest humanoid robotics market report.
This chart, featured in our humanoid robotics market deck, shows how Agility Robotics is capturing share in humanoid robotics
Is UBTECH's humanoid business less dependent on one customer?
UBTECH appears to have a broader order base than Agility, although its latest financial statements reveal that customer concentration has not disappeared.
During its commercial ramp, UBTECH disclosed several separate humanoid contracts rather than one deal carrying nearly the entire order book. Publicly reported agreements included orders worth RMB250 million, RMB159 million, RMB143 million, RMB126 million and RMB264 million across manufacturing and embodied-AI projects.
Industrial relationships have also spread across automotive, electronics, logistics, semiconductor manufacturing and aerospace. UBTECH has worked with companies including BYD, Foxconn and Airbus, while more than 80% of its full-size humanoid applications are now industrial according to the company.
There is still concentration. UBTECH's latest interim accounts say one customer generated RMB308.5 million, or about 24% of total group revenue, during the first half of 2026. That is meaningful exposure to one buyer.
Still, the overall picture differs from Agility's. Agility's biggest disclosed humanoid order can be traced largely to one 1,000-robot agreement. UBTECH has built its humanoid revenue through a wider collection of orders and customers.
That makes UBTECH's commercial base look more mature today, even if we cannot reconstruct an exact current remaining backlog from its disclosures.
Could 1X have more humanoid robots on order than Agility?
1X may already have more customer reservations in unit terms, with 10,000 NEO home robots booked against its planned first year of production, but those reservations are much softer commitments than Agility's industrial contract.
According to 1X, the company filled an entire year of planned NEO production only five days after opening preorders. That means 10,000 people reserved a robot.
The headline value can easily be exaggerated. NEO has a $20,000 Early Access purchase option, so multiplying 10,000 robots by $20,000 gives $200 million. But customers have not committed $200 million.
The reservation requires a $200 deposit, and 1X says that deposit is fully refundable. Buyers may also choose a $499-per-month subscription instead of purchasing the robot outright.
At 10,000 reservations, the deposits would amount to around $2 million if they all followed those terms. The eventual contract value could become dramatically larger if customers convert, but we cannot count that future conversion today.
Still, the 1X number is worth watching. Industrial humanoid companies have generally won demand customer by customer through lengthy enterprise sales cycles. Getting 10,000 consumers to put money down within five days suggests a very different route to scale.
This chart, featured in our humanoid robotics market deck, illustrates yearly funding for humanoid robotics startups
Does Figure really have demand for 100,000 humanoid robots?
Figure has talked about a path to 100,000 robots through major customers, but there is no public evidence of a firm 100,000-unit order book.
The number came from Figure CEO Brett Adcock after the company signed its second major commercial customer. He said the two customer relationships could create a path to shipping 100,000 robots over four years.
“Path to” is doing a lot of work there. Figure has never published signed purchase quantities that add up to 100,000 robots, nor has it disclosed contract values large enough for us to reconstruct such a backlog.
What makes Figure interesting today is that its operating evidence has improved considerably. Figure 02 spent more than 1,250 hours working at BMW's Spartanburg plant, handled more than 90,000 parts and contributed to production of more than 30,000 BMW X3 vehicles.
Figure has since returned to BMW with Figure 03 for a more complicated logistics workflow. The company also reported manufacturing more than 350 Figure 03 robots and increasing production from roughly one robot per day to one per hour in under four months.
Those are meaningful advances in deployment and manufacturing. They still do not give Figure a larger disclosed order book than Agility.
If you want more recent data on this point, please see our latest humanoid robotics market report.
Why don't AgiBot and Unitree lead the order-book ranking if they ship thousands of humanoids?
AgiBot and Unitree currently dominate global humanoid shipments, but neither gives us enough public backlog information to claim the industry's largest order book.
The scale has changed very quickly. Smart Analytics Global estimated that around 19,100 humanoid robots shipped worldwide in the first half of 2026, up roughly 272% from the same period a year earlier.
AgiBot accounted for about 8,400 units, giving it 44% of estimated global shipments. Unitree followed with roughly 5,900 units and 31%. Together they represented about three quarters of the entire market.
Those figures dwarf the shipment volumes of most Western humanoid developers. They also show why an order-book ranking alone can give a strange picture of the industry.
A company selling standardized robots to universities, developers, entertainment customers or smaller commercial users can turn orders into shipments quickly, leaving relatively little backlog at any one moment. Another company can sign a three-year industrial service contract before manufacturing most of the robots involved.
There is also a product-mix issue. AgiBot's shipment figures include several robot families, while Unitree sells the G1 into research, education and other relatively accessible markets. Agility's contract concerns a higher-value enterprise RaaS deployment with recurring services attached.
So AgiBot currently leads the shipment race by a huge margin over Agility. We simply do not have a reliable figure showing how many unfilled orders AgiBot or Unitree still hold.
This chart, featured in our humanoid robotics market deck, compares the main business model options for humanoid robot manufacturers
Is Apptronik sitting on a big hidden order book?
Apptronik probably has a meaningful commercial pipeline, but there is currently no public number that allows us to rank its Apollo humanoid order book against Agility or UBTECH.
Its customer list is serious. Apptronik has announced work with Mercedes-Benz, GXO Logistics and Jabil, and the company has raised more than $900 million to expand Apollo development and manufacturing.
Mercedes-Benz began testing Apollo for manufacturing tasks, while GXO provides Apptronik with exposure to large logistics operations. Jabil is especially interesting because it can act both as a manufacturing partner and a deployment environment.
But none of those relationships currently gives us the metric we need here. Apptronik has not disclosed an aggregate dollar backlog or a firm number of Apollo robots ordered across its customers.
This is common across the sector. A recognizable automotive or logistics customer makes a robotics company look commercially advanced, but a pilot involving a handful of robots and a binding commitment for hundreds of robots are completely different businesses.
Until Apptronik starts publishing quantities or contract values, we would be guessing.
Should Tesla Optimus count as having an enormous order book?
Tesla Optimus should currently be excluded from an external humanoid order-book ranking because Tesla's planned internal deployments are not customer orders.
Tesla has discussed eventually producing Optimus at enormous scale, and Elon Musk has repeatedly described factories as an early use case. Tesla can potentially deploy robots inside its own manufacturing operations before selling large numbers to outside companies.
That is a real strategic advantage. Tesla controls both the robot supplier and the initial factory customer, so it can generate operating data and prove applications without persuading a separate company to sign a contract first.
But an internal plan to manufacture 10,000, 100,000 or eventually one million robots does not create external backlog. No independent buyer has committed capital simply because Tesla decides to use Optimus inside Tesla factories.
For order-book purposes, we therefore keep Tesla outside the ranking until third-party customer commitments become public.
This chart, featured in our humanoid robotics market deck, shows the revenue mix across customer segments in the humanoid robotics market
Who is actually leading humanoid robotics commercially today?
There is no single commercial leader because Agility, UBTECH, AgiBot and Figure lead on different things, but UBTECH has the strongest all-around commercial evidence we can see.
Agility leads the specific metric behind this article: disclosed humanoid-specific contracted value. Its $300 million-plus Digit v5 book is the clearest large backlog we found.
AgiBot leads physical shipment volume. Its estimated 8,400 humanoid shipments in the first half of 2026 represented around 44% of worldwide shipments, according to Smart Analytics Global.
Figure has one of the strongest Western factory-validation stories. Its BMW deployment ran for more than 1,250 hours, handled over 90,000 parts and contributed to production of tens of thousands of vehicles, while Figure 03 manufacturing is now ramping quickly.
UBTECH sits closer to the intersection of all three. It has disclosed large orders, sold more than a thousand full-size humanoids in 2025, sold another 921 in the first half of 2026 and generated RMB590.3 million of half-year revenue from that category.
So we separate “biggest order book” from “winning commercially.” Agility gets the first title. UBTECH currently has the stronger case for the second.
| Metric | Company that stands out now | Best public evidence |
|---|---|---|
| Humanoid-specific contracted value | Agility Robotics | $300M+ Digit v5 orders |
| Broader robotics orders + pipeline | NEURA Robotics | $1B+ across its portfolio |
| Full-size humanoid revenue | UBTECH | RMB590.3M in H1 2026 |
| Humanoid shipment volume | AgiBot | About 8,400 in H1 2026 |
| Consumer reservations | 1X | 10,000 NEO reservations |
| Proven Western factory deployment | Figure / Agility | BMW and multi-customer industrial deployments |
If you want more recent data on this point, please see our latest humanoid robotics market report.
Which humanoid robot company's orders look most trustworthy?
UBTECH currently gives us the strongest evidence that large humanoid orders can repeatedly become deliveries and revenue, while Agility has the largest clearly disclosed future contract.
That difference is more useful than simply comparing two headline numbers.
Agility's main contract is huge relative to the size of its historical business. The company now has to manufacture and deploy 1,000 Digit v5 robots, satisfy contractual milestones and make the economics work over a three-year service period. If it succeeds, Agility's revenue profile could change completely.
UBTECH has already crossed part of that bridge. Full-size humanoid revenue rose from around RMB35.6 million in 2024 to RMB820 million in 2025. Then, rather than slowing sharply, it generated another RMB590.3 million during the first six months of 2026.
As we saw above, those six-month sales reached 921 full-size robots. That makes UBTECH's commercialization easier to trust because we can see orders, deliveries and recognized revenue moving together.
NEURA's $1 billion-plus figure remains too mixed to analyze with the same precision. Figure and Apptronik disclose too little contract data. 1X still needs to show that refundable reservations convert into paying households.
If our goal is predicting who will turn today's demand into a large recurring humanoid business, we would currently put more weight on UBTECH's repeated conversions than on any single backlog announcement.
This chart, featured in our humanoid robotics market deck, shows how factory humanoid robot technology has evolved over time
Which humanoid robotics company really has the biggest order book?
Agility Robotics has the biggest clearly disclosed humanoid-specific order book today, with more than $300 million of contracted Digit v5 orders, so it is our answer to the title.
NEURA Robotics has the larger headline commercial number at more than $1 billion, but its disclosure combines an order book and strategic deployment pipeline across a much broader robotics portfolio. We cannot credibly present that as $1 billion of humanoid backlog.
UBTECH comes next with RMB1.4 billion, roughly $200 million, of humanoid orders reported for 2025. What makes UBTECH particularly important is what happened afterward: its full-size humanoid business generated RMB820 million in 2025 revenue and another RMB590.3 million in the first half of 2026. The company is showing that humanoid demand can turn into meaningful sales repeatedly rather than remaining on an order sheet.
The Chinese volume leaders complicate the picture further. AgiBot shipped roughly 8,400 humanoids in only six months, while Unitree shipped about 5,900. Their undisclosed remaining backlogs could be substantial, but there is no defensible public number that lets us put either above Agility.
Figure's potential path toward 100,000 robots belongs in the same category as 1X's 10,000 refundable reservations and Tesla's enormous Optimus production ambitions: important evidence about possible future scale, but weaker evidence about committed revenue today.
So the clearest answer is Agility Robotics. The more interesting conclusion is that Agility's lead is narrower than the $300 million headline makes it look. UBTECH already has a broader stream of humanoid revenue, AgiBot is shipping robots at a completely different scale, and several major competitors disclose almost no backlog at all. The company with the biggest visible order book today may therefore be easier to identify than the company that ultimately sells the most humanoid robots.
OUR METHODOLOGY
This analysis treats the question as a structured evidence problem rather than a comparison of the largest numbers humanoid-robot companies have announced. We looked at the strength of the customer commitment, how specifically each disclosure applies to humanoids, whether demand is still forward-looking or has already converted into deliveries and revenue, evidence of real-world deployment, and the quality and recency of the underlying disclosure.
We gave the most weight to signed purchase agreements, Robotics-as-a-Service contracts and explicitly reported humanoid procurement orders. Conditional frameworks, refundable reservations, sales pipelines, possible future volumes and production targets were kept separate because they do not represent the same level of committed demand.
We did not force unlike disclosures into one score. Agility's contracted Digit orders, UBTECH's annual humanoid order total, NEURA's broader order book plus strategic deployment pipeline, 1X's refundable reservations and Figure's prospective customer volumes are useful in different ways, but they are not interchangeable.
We also checked whether large demand claims were reinforced by operating evidence. Agility's warehouse deployments and 65,000-plus operating hours, UBTECH's recognized humanoid revenue and unit sales, Figure's BMW deployment, and AgiBot and Unitree shipment estimates help distinguish a backlog story from a business that is already delivering robots at scale.
For Agility Robotics, the key sources were its latest SEC S-4 registration statement, its SEC-filed investor presentation, the SEC-filed transaction announcement, GXO's multi-year Digit deployment announcement, and Schaeffler's 2026 AGM report.
For NEURA Robotics, we used its 2026 Series C announcement, its 2025 Series B announcement, and its ek robotics acquisition announcement to distinguish the company's broad robotics demand pool from humanoid-specific backlog.
For UBTECH, the main sources were its 2025 annual results filed with HKEX, its H1 2026 interim results filed with HKEX, and Xinhua's reporting on 2025 humanoid orders and the Airbus relationship.
Other key sources included 1X's NEO Factory update and NEO order page; Brett Adcock's statement on Figure's path toward 100,000 robots; BMW Group's Figure deployment update; Figure's Figure 03 manufacturing-ramp update and production strategy; Apptronik's expanded Series A announcement and Jabil collaboration; Tesla's Q2 2026 investor update; and Smart Analytics Global's H1 2026 humanoid shipment estimates.
The final judgment comes from aggregating the strongest recent evidence across these dimensions while keeping related questions separate. Having the largest visible order book, shipping the most robots and showing the strongest commercialization are not necessarily the same thing, which is why Agility can lead the title metric while UBTECH looks stronger on repeated conversion into revenue.
In our humanoid robotics market deck, we identify pain points entrepreneurs should prioritize
Related blog posts
- Which humanoid robotics companies are actually making money?
- Which humanoid robotics startups are actually making the most money today?
