Humanoid robot sales reached about 7,000 worldwide in 2025, according to the International Federation of Robotics' first industry-wide count of the emerging category. The figure establishes a baseline for a market attracting billions of dollars but still operating mainly through research purchases and small factory pilots.

 

The IFR data, reviewed by Reuters before the federation publishes its wider 2025 robot statistics on September 24, cover machines sold for industrial and professional-service use. They exclude consumer and military humanoids, while medical robots are classified separately.

 

What the Humanoid Robot Sales Count Includes

The federation defines a humanoid as an autonomous robot with a human-like appearance that can operate in environments designed for people. The definition does not require legs, allowing wheeled or stationary machines to qualify when their form and behavior otherwise meet the criteria.

 

That scope matters because manufacturers use the label inconsistently. Some products are full bipedal machines with arms and hands; others place a human-shaped torso on a mobile base. IFR's methodology begins turning those varied designs into a measurable industry category.

 

The inaugural tally has four important boundaries:

  • About 7,000 units were sold during 2025
  • Industrial and professional-service robots are included
  • Consumer and military machines are excluded
  • Medical robots remain a separate classification

 

The figures come from robot suppliers and national industry associations. IFR has collected industrial and service-robot statistics for roughly three decades through its World Robotics reports, but this is the first year it has compiled humanoid data alongside those established markets.

 

Most 2025 Buyers Were Still Testing the Technology

A large share of the humanoids sold last year were not assigned to productive shifts. IFR Secretary General Susanne Bieller told Reuters that research institutions and companies bought many units to collect data and improve artificial-intelligence models.

 

Automakers are among the most visible early adopters, using humanoids for material handling, repetitive assembly support and tests in human-designed workspaces. Yet current deployments generally involve single-digit or low double-digit numbers of machines at individual plants, according to the federation.

 

This pattern separates commercial sales from commercial maturity. A manufacturer can book revenue when a laboratory or corporate innovation team buys a robot, even if the machine has not demonstrated reliable, economical work in continuous production.

 

The 7,000-unit total is therefore best read as the size of the development ecosystem. It measures how many machines entered industrial and research settings, not how many human jobs were replaced or how many factories now depend on general-purpose humanoids.

 

Humanoids Remain Tiny Beside Established Robot Markets

For scale, approximately 542,000 conventional industrial robots were installed in 2024. Another estimated 199,000 professional service robots were sold for uses such as transport, hospitality and cleaning during that year. The humanoid count remains a small fraction of either established category.

 

Traditional industrial arms benefit from decades of engineering, standardized safety practices and clear returns on investment. Service robots often handle a narrow task in a controlled setting. Humanoids promise flexibility across many human environments, but that broader ambition brings harder problems in balance, dexterity, battery life, perception and safety.

 

China is central to the industry's growth expectations. The country is already the largest market for conventional industrial robots and has made embodied intelligence a technology priority. Chinese manufacturers are expanding production capacity while state programs, research centers and factories create demand for pilot units.

 

Forecasts Run Far Ahead of the IFR Baseline

Bank of America Global Research estimates that 90,000 humanoid robots could ship in 2026 and 1.2 million in 2030. Those are forecasts from a financial institution, not IFR sales projections, and they imply an extremely steep expansion from the measured 2025 base.

 

Reaching that trajectory would require more than scaling factories. Vendors must prove that robots can complete useful tasks for long periods, recover from errors, work safely near people and deliver a lower total cost than purpose-built automation or human labor.

 

The new IFR category should make those claims easier to test. Annual counts can show whether announced production plans convert into shipments, whether pilots grow into fleets and which regions or applications create sustained demand.

 

A Baseline for Measuring Humanoid Deployment

The industry's promotional videos have advanced faster than its standardized statistics. IFR's first tally narrows that information gap by providing a consistent definition and a supplier-backed estimate, even though the initial total remains small.

 

The next signal will be repeat deployment. If manufacturers return after pilots to order dozens or hundreds of units for defined jobs, the market will begin moving from development hardware toward operational automation. If purchases remain concentrated in research and data collection, growth forecasts will need a longer timetable.

 

For now, 7,000 is both a milestone and a reality check. Humanoid robots have become a trackable commercial market, but the first global count shows that widespread productive use is still at an early stage.

 

Related Research