A Humanoid Analytics review of its expanded company tracker identifies 225 companies and major corporate programs classified as core humanoid robot builders across 21 countries. China accounts for 144 of them, or 64% of the tracked total, followed by the United States with 31. South Korea and Japan are tied for third place with seven core builders each.
The charts show a humanoid robotics sector that is substantially larger and more geographically concentrated than the most visible group of well-funded Western startups might suggest. They do not, however, show that 225 commercially ready humanoid robot companies exist.
That distinction is important. The Humanoid Company Tracker is a curated research database covering robot developers, major corporate programs, enabling technology suppliers and selected ecosystem organizations. It classifies entities by scope, geography, segment, public status, evidence stage, commercial readiness and confidence. Inclusion does not mean that a company has a commercially mature robot or a verified customer deployment. The Market Is Larger Than the Builder Count
The first chart separates the tracker’s 359 records into four groups.
Core humanoid builders account for 225 entries, or approximately 62.7% of the tracker. These include companies directly developing humanoid platforms as well as major corporate humanoid programs.
A further 126 entries are classified as adjacent technology companies. This group includes businesses working on dexterous manipulation, tactile sensing, robot components, control systems, embodied intelligence, data platforms and other technologies relevant to humanoid development. Six records represent research, innovation or ecosystem organizations, while two are excluded or non-qualifying entries.
This separation prevents a common market-counting error. A company that produces a robotic hand, force sensor, actuator or embodied AI model may be important to the humanoid supply chain, but it should not automatically be counted as a humanoid robot manufacturer.
The difference between 359 tracked entities and 225 core builders is therefore not a minor methodological detail. It changes how the market should be understood. The larger figure describes the visible ecosystem. The smaller figure is the more defensible count of companies and corporate programs directly associated with developing humanoid platforms.
Even the 225 figure should be described as a tracker snapshot rather than a definitive global census. Some entities have detailed official product information, while others remain directory-listed or supported by incomplete public evidence. The tracker marks directory-only records as low confidence until stronger information becomes available. China Dominates the Tracked Company Population
The geographic chart shows the strongest finding in the dataset: China represents 144 of the 225 core builders, equal to 64%.
The United States is a distant second with 31 companies, or 13.8%. South Korea and Japan each account for seven, or 3.1%. The remaining 36 companies are distributed across 17 countries and collectively represent 16% of the tracked builder population.
This concentration is commercially significant. A large builder population can increase competition for engineering talent, components, manufacturing partners, customers and government support. It can also encourage faster product iteration and create demand for common supply-chain inputs such as actuators, motors, reducers, batteries, sensors and dexterous hands.
However, company count is not the same as industrial leadership.
The Chinese total includes companies at very different stages. Some have public products, shipment claims, funding and announced customer activity. Others remain prototype developers, corporate research programs or low-confidence directory entries whose current products and operating status still require verification. The tracker itself repeatedly distinguishes product announcements and shipment claims from customer-level deployment proof. same caution applies to the United States. Its smaller company count does not measure the amount of capital available, the technical quality of individual platforms or the strength of named customer programs. A country with fewer builders could still contain companies with stronger funding, manufacturing infrastructure or deployment evidence.
The chart therefore measures market participation and visible company formation. It does not rank countries by robot performance, deployed fleets, revenue, production output or commercial readiness.
South Korea and Japan Share Third Place
South Korea and Japan each have seven core entries, placing them jointly behind China and the United States.
The tie is analytically useful because it shows why country counts should not be interpreted without examining the underlying organizations. The South Korean group includes standalone robot developers and major corporate programs, while the Japanese group includes commercial platforms, corporate research programs and long-running humanoid development efforts. Individual entries also differ significantly in product type, maturity and intended use. equal number of tracked builders does not establish equal commercial capability. A productized upper-body research platform, a bipedal prototype and a corporate humanoid research program may all qualify for inclusion while representing very different market positions.
The next useful comparison should therefore move beyond company count. It should examine how many builders in each country have identifiable products, named customer pilots, confirmed operating deployments, disclosed production, repeat orders or measurable performance data.
The More Important Market Is Still Much Smaller
The charts demonstrate that humanoid robotics has become a broad international field rather than a contest involving only a few famous companies. They also reveal a large enabling ecosystem operating alongside the robot builders.
What the charts cannot establish is how many of the 225 core builders will become sustainable commercial vendors.
That number is almost certainly smaller than the headline count because many tracked organizations remain at the prototype, demonstration, internal-testing or early-product stage. The decisive evidence will come from named customer operation, payment, repeated task performance, deployment duration, uptime, intervention rates, safety results, fleet expansion and repeat purchasing.
For investors and suppliers, the 225-company universe indicates a large opportunity pipeline but also a high risk of overcounting credible customers. For industrial buyers, it shows that vendor choice is expanding while reliable deployment evidence remains uneven. For competitors, China’s 64% share signals intense company formation, but not yet a settled commercial outcome.
The assessment would change when more low-confidence entries are independently verified or removed, and when country-level counts can be compared with confirmed production and customer deployment evidence. Until then, the most accurate conclusion is that Humanoid Analytics tracks 225 core humanoid builders, not that the world has exactly 225 commercially viable humanoid robotics companies.
Sources:
Humanoid Analytics, “Humanoid Company Tracker”: https://humanoidanalytics.com/humanoid-company-tracker/