Skip to content
humanoidrobots.training

Sim benchmark · 15 models · policies · 0 submitted

Leaderboard

Full benchmark with method, sortable columns and per-model pages: bench.humanoidrobots.training

Two boards. The sim benchmark runs every published model headless each night under one generic controller, so a failure here is a fact about the model and our controller, not a verdict on the robot. The policy board ranks trained policies once they arrive with an eval we can rerun on the same pinned MJCF.

Nightly stand-hold benchmark

11/15 hold a stand · run 2026-09-24 · MuJoCo 3.14.0 WASM

Stand-hold benchmark per simulation model. Passing models first, then alphabetical.
RobotModel commitStand-hold 10 sBase driftRTFLoadDateMethod
Apptronik ApolloApptronikc96a32dmujoco_menageriepassCoM min 99% · tilt 1.5°-1.0 cm55.5×2.98 s2026-09-24stand-hold-v1
Berkeley HumanoidUC Berkeley (Hybrid Robotics)c96a32dmujoco_menageriepassCoM min 98% · tilt 4.5°-1.2 cm49.0×1.30 s2026-09-24stand-hold-v1
Booster T1Booster Roboticsc96a32dmujoco_menageriepassCoM min 99% · tilt 4.9°0.0 cm29.5×0.41 s2026-09-24stand-hold-v1
Booster T1 12-DoFBooster Roboticsda396a0booster_gympassCoM min 95% · tilt 0.3°-2.7 cm51.8×0.39 s2026-09-24stand-hold-v1
Fourier N1Fourierc96a32dmujoco_menageriepassCoM min 100% · tilt 0.7°-0.1 cm38.2×2.69 s2026-09-24stand-hold-v1
PAL TALOSPAL Roboticsc96a32dmujoco_menageriepassCoM min 99% · tilt 1.9°-0.3 cm6.0×0.72 s2026-09-24stand-hold-v1
ROBOTIS OP3ROBOTISc96a32dmujoco_menageriepassCoM min 93% · tilt 0.6°-2.1 cm33.6×1.87 s2026-09-24stand-hold-v1
Unitree G1Unitree Roboticsc96a32dmujoco_menageriepassCoM min 100% · tilt 0°+0.2 cm26.4×1.09 s2026-09-24stand-hold-v1
Unitree G1 12-DoFUnitree Robotics276801eunitree_rl_gympassCoM min 100% · tilt 0.8°-0.2 cm42.2×1.14 s2026-09-24stand-hold-v1
Unitree H1Unitree Roboticsc96a32dmujoco_menageriepassCoM min 100% · tilt 0.7°-0.4 cm41.9×0.56 s2026-09-24stand-hold-v1
Unitree H1 10-DoFUnitree Robotics276801eunitree_rl_gympassCoM min 99% · tilt 0.8°-1.4 cm56.0×0.71 s2026-09-24stand-hold-v1
Agility CassieAgility Roboticsc96a32dmujoco_menageriefellCoM min 17% · tilt 88.3°-84.8 cm8.9×0.46 s2026-09-24stand-hold-v1
DropbearHyperspawnfloor-only contact · meshes simplifiedafe0ea0dropbear_mjcffellCoM min 13% · tilt 103°+23.3 cm0.6×8.14 s2026-09-24stand-hold-v1
PNDbotics Adam LitePNDboticsc96a32dmujoco_menageriefellCoM min 11% · tilt 90.1°-83.0 cm55.1×4.13 s2026-09-24stand-hold-v1
ToddlerBot 2XCStanfordc96a32dmujoco_menageriefellCoM min 32% · tilt 149.8°-19.8 cm15.9×2.52 s2026-09-24stand-hold-v1
Method
stand-hold-v1
Controller
The browser sim's hold controller (packages/sim/src/engine.ts), mirrored headless: joint PD toward the stand keyframe (or qpos0) plus an ankle-strategy balance term from torso tilt, per-model gains from the @hs/sim manifest. No policy, no harness, no pushes.
Pass
Runs the full 10 s without a physics reset, the centre of mass never drops below 80% of its start height, and it ends within 10% of the start. Base drift is reported alongside.
Fell
The centre of mass drops below 60% of its start height at any point, or the physics diverges (NaN / auto-reset). Heights use the whole-body centre of mass because some models put the base origin near the floor.
RTF
Simulated seconds divided by wall seconds for the 10 s run, single thread, no rendering. Depends on the machine; see environment.
Load
Compile time: writing the pinned files into the WASM filesystem, mesh simplification when the model's runtime says so, MjModel.from_xml_path, and controller setup. Download time is excluded.
Machine
local · 13th Gen Intel(R) Core(TM) i9-13980HX ×32 · win32 10.0.26200 x64 · Node v23.11.0
Data
Written nightly by the Coach agent (engine) to apps/humanoidrobots.training/data/benchmarks.json and merged through a pull request.

Policy board

  • Walk 0
  • Walk, rough terrain 0
  • Push recovery 0
  • Motion imitation (dance) 0
#PolicyRobotTrainerEval returnSim stepsRan on hardware

No policies yet. Zero is the real number.

Train one with mjlab and send the checkpoint, its eval video and the task id. The first entry goes here.

Set up a run