Robot 500

Issue 01 · October · Published 2 October 2026

The Robot 500

The monthly ranking of advanced robots. Humanoids, robot dogs, cobots, drones, service robots and air taxis, ranked like for like on published evidence.

  • 76 ranked
  • 6 categories
  • 65 on the watchlist
  • 0 paid placements
Unitree H1 Humanoid robots · joint number one=1Unitree H1Unitree Robotics · score 55.5 · grade B
01

Humanoid robots

29 ranked · leader: Unitree H1

All 29 humanoid robots →

02

Robot dogs

14 ranked · leader: Unitree B2

All 14 robot dogs →

03

Cobots

13 ranked · leader: ABB GoFa (CRB 15000)

All 13 cobots (collaborative robot arms) →

04

Drones

16 ranked · leader: DJI Mavic 4 Pro

All 16 camera drones →

05

Service robots

2 ranked · leader: KEENON DINERBOT T9

Fewer than 5 ranked, so scores are shown without positions.

06

Air taxis (eVTOL)

2 ranked · leader: Joby S4

Fewer than 5 ranked, so scores are shown without positions.

Questions

What is the best humanoid robot?

In the Robot 500 October ranking, the Unitree H1 leads the humanoid robots (score 55.5/100, evidence grade B), followed by EngineAI T800, Unitree G1, 1X NEO. Ranks are based on published, sourced evidence, not hands-on tests.

What is the best robot dog?

The Unitree B2 leads the robot dogs (quadrupeds) in the Robot 500 (score 64.1/100), ahead of Unitree B2-W and DEEP X30.

What is the Robot 500?

A monthly, independent ranking of advanced robots (humanoids, robot dogs, cobots, drones, service robots and air taxis), ranked like for like within each category on published evidence: 70% sourced performance figures and 30% real-world proof. It is published by RankedRobot and is never for sale.

How often is the Robot 500 updated?

Once a month, on the 1st. The current ranking was published on 2 October 2026; the next update is on 1 November 2026.

Can a company pay to rank higher?

No. There are no paid placements and no manual overrides; a robot moves only when its published evidence changes.

How to read the Robot 500

Robots are compared only with robots of the same kind. Score = 70% performance (sourced specifications against category rivals) + 30% real-world proof (on sale, deployments, certification, durability rating, independent data). Grades A–C show how much evidence stands behind each rank. Scores less than 1 point apart share a position (=2). Price never affects a rank. Next update: 1 November 2026. The full method →