Humanoid Robot Smashes 400m World Record at Beijing Games

The World Humanoid Robot Games in Beijing saw machines surpass human athletic achievements this past weekend, with one robot setting a new 400-meter world record.

Tiangong Ultra, a white humanoid robot developed by a Beijing-based research center, completed the 400-meter sprint in 38.15 seconds, shattering the existing human record of 43.03 seconds set by Wayde van Niekirk in 2016. This feat followed Tiangong Ultra's performance on Saturday, where it reportedly broke Usain Bolt's 100-meter world record during the games' opening ceremony. However, its 100-meter victory was marred by an incident where the robot failed to decelerate and collided with a safety mattress, requiring it to be carried off for repairs.

Games Overview

The five-day event, held in Beijing's Olympic speed skating arena, is only the second installment of the World Humanoid Robot Games. This year's competition features 2,056 robots registered by 666 teams, with 96 percent of teams originating from China. Participants from 16 countries across six continents are competing in 1,301 matches. The games include a mix of sports-inspired events such as athletics, boxing, football, and martial arts, alongside task-based challenges that showcase potential commercial applications for humanoid robots.

Of the 50 scheduled events, 30 are sports competitions, including weightlifting, gymnastics, football, tug of war, and 11 Olympic-style track and field events. The remaining 20 events focus on work scenarios, testing capabilities in areas like bricklaying, power tool assembly, and package handling. Autonomous robots are awarded full points, while semi-autonomous bots receive half points. Notably, the 2026 games do not include events featuring stairs, ladders, uneven terrain, or human-robot collaboration.

Audience reactions included amusement, particularly when a boxing robot stumbled and fell out of the ring. One spectator, Hou Yali, expressed interest in seeing robots compete against humans in more complex tasks, such as household assistance, noting that current competitions focus solely on speed and task execution rather than broader intelligence or practical application.