The event has also expanded from a three-day format with six competition units to five days and nine units, according to reporting from the organizing committee.
The games will include teams from 16 countries across six continents. The reports specifically identify participation by robotics powers including the United States, Germany and Japan, alongside China and Brazil. Brazil is assembling five RoboCup teams into a national squad for the event.
The supplied reports do not provide a complete official list of all 16 countries, so the named countries should not be treated as the full roster.
China is expected to account for most of the field. One report describes 641 Chinese teams and 1,975 robots representing companies, universities and research institutes. More broadly, domestic participation includes robotics enterprises as well as teams from higher-education and research organizations.
The clearest increase is in team participation. The second edition has 666 registered teams, compared with 280 teams at the inaugural games—an increase of 138%.
The robot count has reached 2,056. Organizers and several reports characterize that as approximately four times the inaugural turnout, although another supplied report describes the increase differently. The safest conclusion is that robot participation has risen sharply, while the available reports do not agree on a single precise multiplier.
The scale of competition has grown as well. Earlier announcements described more than 30 events, but the final published programme lists 51 events and 1,301 sessions.
The competitive programme adds long jump, weightlifting, tug of war and table tennis. These events are designed to expose weaknesses in balance, force control, coordination, mechanical design and core components—not simply whether a robot can walk or run.
The programme also includes football, combat, dance and track-and-field events. The games therefore function as both a public competition and a repeatable engineering test: a small improvement in jumping distance or lifting capacity can reflect advances in motors, reducers, joints, control systems and precision manufacturing.
The most consequential change is the emphasis on scenario-based events. Rather than testing robots only on a sports field, the games simulate environments such as:
Tasks include industrial assembly, housekeeping, emergency response and sorting books. A dexterous-hand challenge reportedly tests operations such as tool assembly, powder weighing, block stacking, nail driving, opening bottles, unpacking items, picking up objects with tweezers and connecting cables.
These tasks focus on the difficult final step between a robot recognizing an object and reliably manipulating it in an unpredictable, human-oriented setting. That is why the organizers emphasize fine, “last-meter” capabilities and increasingly autonomous performance.
Organizers present the competition as an evolving testbed for embodied intelligence and practical deployment. Their stated expectation is that tougher rules, more autonomous events and better fine manipulation will help convert laboratory and competition advances into industrial systems and workplace orders.
That goal is not the same as evidence that widespread commercialization has already happened. The 2026 games can demonstrate progress under defined conditions, but the ability to complete a contest does not by itself prove that a robot is ready for dependable, cost-effective work across ordinary industrial or service environments.
The strongest signal from this year’s event is therefore not any single race or match. It is the combination of larger participation, harder physical tests and practical scenarios. Together, those changes show how humanoid robotics competitions are being shaped to measure not only movement, but also whether machines can perform useful tasks with increasing autonomy.
The event has also expanded from a three-day format with six competition units to five days and nine units, according to reporting from the organizing committee.
The games will include teams from 16 countries across six continents. The reports specifically identify participation by robotics powers including the United States, Germany and Japan, alongside China and Brazil. Brazil is assembling five RoboCup teams into a national squad for the event.
The supplied reports do not provide a complete official list of all 16 countries, so the named countries should not be treated as the full roster.
China is expected to account for most of the field. One report describes 641 Chinese teams and 1,975 robots representing companies, universities and research institutes. More broadly, domestic participation includes robotics enterprises as well as teams from higher-education and research organizations.
The clearest increase is in team participation. The second edition has 666 registered teams, compared with 280 teams at the inaugural games—an increase of 138%.
The robot count has reached 2,056. Organizers and several reports characterize that as approximately four times the inaugural turnout, although another supplied report describes the increase differently. The safest conclusion is that robot participation has risen sharply, while the available reports do not agree on a single precise multiplier.
The scale of competition has grown as well. Earlier announcements described more than 30 events, but the final published programme lists 51 events and 1,301 sessions.
The competitive programme adds long jump, weightlifting, tug of war and table tennis. These events are designed to expose weaknesses in balance, force control, coordination, mechanical design and core components—not simply whether a robot can walk or run.
The programme also includes football, combat, dance and track-and-field events. The games therefore function as both a public competition and a repeatable engineering test: a small improvement in jumping distance or lifting capacity can reflect advances in motors, reducers, joints, control systems and precision manufacturing.
The most consequential change is the emphasis on scenario-based events. Rather than testing robots only on a sports field, the games simulate environments such as:
Tasks include industrial assembly, housekeeping, emergency response and sorting books. A dexterous-hand challenge reportedly tests operations such as tool assembly, powder weighing, block stacking, nail driving, opening bottles, unpacking items, picking up objects with tweezers and connecting cables.
These tasks focus on the difficult final step between a robot recognizing an object and reliably manipulating it in an unpredictable, human-oriented setting. That is why the organizers emphasize fine, “last-meter” capabilities and increasingly autonomous performance.
Organizers present the competition as an evolving testbed for embodied intelligence and practical deployment. Their stated expectation is that tougher rules, more autonomous events and better fine manipulation will help convert laboratory and competition advances into industrial systems and workplace orders.
That goal is not the same as evidence that widespread commercialization has already happened. The 2026 games can demonstrate progress under defined conditions, but the ability to complete a contest does not by itself prove that a robot is ready for dependable, cost-effective work across ordinary industrial or service environments.
The strongest signal from this year’s event is therefore not any single race or match. It is the combination of larger participation, harder physical tests and practical scenarios. Together, those changes show how humanoid robotics competitions are being shaped to measure not only movement, but also whether machines can perform useful tasks with increasing autonomy.