Half marathon: machine 50:26 against human 56:51
A humanoid robot covered the half marathon in 50:26 at the Beijing E-Town Humanoid Robot Half Marathon in April 2026, faster than Yomif Kejelcha’s 56:51 human world record. The machine ran in a humanoid division that permits battery swaps and handler contact, so Overtaq records the crossing as qualified rather than confirmed.
The record in full
| Discipline | Half marathon ATHLETICS · ROAD |
|---|---|
| What is measured | Time over 21.0975 km on a road course |
| Human benchmark | 56:51 — Yomif Kejelcha, Ethiopia ET · 23 AUGUST 2026 · MEDIA MARATÓN CIUDAD DE BUENOS AIRES |
| Machine benchmark | 50:26 — Lightning 19 APRIL 2026 · BEIJING E-TOWN HUMANOID ROBOT HALF MARATHON |
| Attributed to | Honor Robot CHINA |
| Difference | Machine ahead by 6.42 (11.3%) |
| Crossover | 19 April 2026 — Beijing E-Town Humanoid Robot Half Marathon A machine covered the half marathon distance in 50:26, inside every ratified human world record for the distance, before or since. The mark standing today is Yomif Kejelcha’s 56:51 from August 2026. |
| Verification status | REPORTED Independently reported |
| Comparability status | MODIFIED Modified conditions |
| Overtaq classification | QUALIFIED The machine result is independently reported, but the conditions differ materially from the human benchmark, so it is not counted as a confirmed crossover. |
| Data last checked | 23 AUGUST 2026 |
Conditions and comparability
Machine conditions
Ran a humanoid division alongside the human race. Battery changes and human handlers are permitted in the humanoid category; the human record allows neither.
Why this is modified conditions
This is the clearest machine speed advantage in the dataset and the least like-for-like race. Humanoid entrants are allowed battery swaps, accompanying operators and mid-race intervention. Strip those out and there is no comparable result. Overtaq publishes the number and the asterisk together.
Machine progression
2 recorded results. Overtaq keeps every published result rather than only the current best, so the frontier can be read as a history. This table is the full data behind the chart on the interactive record.
| Date | Result | System | Status | Note |
|---|---|---|---|---|
| 19 April 2025 | 2:40:42 | Tiangong Ultra X-Humanoid / Beijing Innovation Center |
REPORTED MODIFIED | Won the first humanoid half marathon, in Beijing’s Yizhuang district, with battery swaps. |
| 19 April 2026 | 50:26 | Lightning Honor Robot |
REPORTED MODIFIED | Cut the machine frontier by more than three quarters in twelve months, and passed the human world record in the process. |
Evidence
3 sources. Every source is a direct link to the primary record where one exists; where Overtaq has not been able to confirm an exact permanent URL, the entry says so rather than linking to a search page.
- World Athletics — FLASH: Kejelcha regains world half marathon record with 56:51 in Buenos Aires
23 AUGUST 2026
Human benchmark of 56:51, Yomif Kejelcha, Media Maratón Ciudad de Buenos Aires, 23 August 2026, subject to ratification.
- Reuters — Humanoid robots run alongside humans in Beijing half marathon
19 APRIL 2025
Tiangong Ultra finishing the first humanoid half marathon in 2:40:42, and the battery-swap rules.
NO EXACT URL — NEEDS MANUAL VERIFICATION
No exact article URL located for the April 2025 Yizhuang humanoid half marathon.
- Al Jazeera — Humanoid robot finishes Beijing half marathon in 50:26
12 APRIL 2026
The 50:26 machine mark and its comparison to elite human pace.
NO EXACT URL — NEEDS MANUAL VERIFICATION
No exact article URL located for the 50:26 half marathon report.
What is still missing
- Rules for the humanoid division — permitted battery swaps, handler contact, course measurement — have not been published in full.
- The 56:51 human mark is subject to the usual World Athletics ratification procedure. Overtaq does not assert which mark was in force on the day of the machine race; the machine time is inside all of them.
Social cards
Every record is rendered into four formats in both themes from the same data, so the verification and comparability status travels with the numbers wherever the card is posted.
