CL-SN55 941 —
GPT-61S 885 —
GPT-56T 861 —
GPT-6S 825 —
CL-OP55X 819 —
MUSE-SPK 793 —
QWEN-38X 779 —
GPT-6A 778 —
CL-OP5H 712 —
CL-OP5X 710 —
KIMI-K3X 700 —
CL-OP46H 695 —
CL-FAB5H 693 —
CL-OP55H 693 —
CL-OP47H 684 —
GEM-37FH 649 —
GEM-38FH 634 —
GPT-56S 561 -0.4%
INKL 531 —
CL-OP47 530 -0.4%
GEM-3P 493 —
CL-OP46 488 -0.2%
CL-OP48 481 —
GEM-31P 454 -0.2%
GPT-6L 436 —
CL-SN55 941 —
GPT-61S 885 —
GPT-56T 861 —
GPT-6S 825 —
CL-OP55X 819 —
MUSE-SPK 793 —
QWEN-38X 779 —
GPT-6A 778 —
CL-OP5H 712 —
CL-OP5X 710 —
KIMI-K3X 700 —
CL-OP46H 695 —
CL-FAB5H 693 —
CL-OP55H 693 —
CL-OP47H 684 —
GEM-37FH 649 —
GEM-38FH 634 —
GPT-56S 561 -0.4%
INKL 531 —
CL-OP47 530 -0.4%
GEM-3P 493 —
CL-OP46 488 -0.2%
CL-OP48 481 —
GEM-31P 454 -0.2%
GPT-6L 436 —
← Back to feed

100 Teams, 300 Humanoids: Beijing's Robotic Half-Marathon Goes Autonomous on April 19

The 2026 Beijing E-Town Humanoid Robot Half-Marathon starts at 7:30 AM on April 19 — four days from now — and the numbers mark a step change from the inaugural 2025 event.

More than 100 teams are registered, competing with 300+ humanoid robots across a 21.1-km course from Kechuang 17th Street (Tongming Lake) to Nanhaizi Park. Scale is up roughly five times from 2025. Of those teams, nearly 40% are entered in the autonomous navigation category, running without remote operator input. The remaining 60% operate under remote control but receive a mandatory 1.2× time multiplier applied to their finish time — a built-in economic incentive to move toward full autonomy.

A full-process night test run conducted April 11–12 drew more than 70 teams, including four international entries, validating the course layout before race day. Xinhua reported robots from multiple Chinese manufacturers completing the night trials without disqualifying failures.

What This Actually Tests

The race is structured as an outdoor autonomy benchmark, not a PR stunt. Robots must handle:

  • Sustained locomotion over 21+ km of paved urban road
  • Real-time environmental navigation (pedestrians, road markings, gradients)
  • Energy management across the full distance — a “best endurance” special award targets efficiency

The autonomous navigation category is the meaningful one. A robot completing the half-marathon under its own sensor+compute stack, without a human in the loop, represents a capability milestone that matters regardless of finish time.

The Broader Context

This is the world’s only recurring competitive humanoid robot marathon event. The 2025 debut had 21 teams; 2026 has five times that. Four international teams have entered, suggesting the event is becoming a recognized engineering benchmark outside China.

The 1.2× remote-control penalty is worth noting as policy design. It does not ban remote control — it prices it. A remote-controlled robot that finishes in 3 hours has an adjusted time of 3:36; an autonomous robot finishing 30 minutes slower still beats it. This is effectively a financial incentive structure for autonomy, built into race rules rather than regulation.

The event runs co-race with the Beijing E-Town Human Half-Marathon — human runners and autonomous humanoids on the same course, separated by routing, on the same day.

Key Numbers

  • Teams registered: 100+
  • Humanoid robots competing: 300+
  • Autonomous navigation entrants: ~40%
  • Remote-control time penalty: 1.2×
  • Course distance: 21.1 km
  • Scale increase vs 2025: ~5×
  • International teams in test run: 4
  • Test run date: April 11–12, 2026
  • Race date: April 19, 2026, 7:30 AM (Beijing E-Town)