Robot library
Every task is pinned to one MuJoCo robot. Start with a tiny robot like the ones on your workbench, learn the workflow on a classic, or bring your own model.
My robot isn’t on this shelfUpload MJCF or URDF, including a ZIP with meshes. Onshape → onshape-to-robot → kualia works well for 3D-printed robots.Desk-sized robots people actually build and buy: open-source bipeds, the LeRobot arms, a nano quadcopter. They train on kualia today.
MODEL 013D-printed desk biped · 16 actuators · ≈35 cm
The open-source, 3D-printed duck-shaped biped inspired by Disney's BDX droids. kualia ships its MJCF with a free-floating base and the Open Duck Playground's standing pose, so you can train walking and balance policies in MuJoCo without writing an environment.
Open Duck project · Apache-2.0
MODEL 02open-source 0.5 m humanoid · 30 actuators · ≈50 cm
Stanford's open-source, 3D-printed humanoid with legs, waist, arms and neck. kualia uses the MuJoCo Menagerie position-controlled model that starts from its home pose.
Stanford · MIT
MODEL 03small hobby humanoid · 20 actuators · ≈50 cm
The 50 cm Dynamixel-powered humanoid used in RoboCup kid-size and university labs, from MuJoCo Menagerie.
ROBOTIS · Apache-2.0
MODEL 04desktop arm + gripper · 6 actuators · ≈40 cm reach
The low-cost desktop arm from Hugging Face LeRobot: five joints and a gripper. Train reach, push and press policies in simulation and export ONNX for the real arm.
TheRobotStudio · LeRobot · Apache-2.0
MODEL 053D-printed arm + gripper · 6 actuators · ≈25 cm reach
The original 3D-printed low-cost arm with Dynamixel servos and a gripper — small, quick to train and forgiving.
Alexander Koch · Apache-2.0
MODEL 062-link servo arm · 2 actuators · ≈36 cm reach
Two Dynamixel servos on a stand — the simplest real arm and a great first reinforcement-learning project; it learns in minutes.
Rhoban · MIT
MODEL 0727 g quadcopter · 4 actuators · 27 g
Bitcraze's palm-sized open-source quadcopter: total thrust plus roll, pitch and yaw moments. A zero action holds the hover trim, so hover and waypoint policies train quickly.
Bitcraze · MIT
Gymnasium’s simplified MuJoCo bodies. Fast to train, ideal for learning the workflow and quick experiments.
MODEL 08cart-pole · 1 actuators
The classic cart-pole from the Gymnasium MuJoCo suite — the gentlest first robot for learning the workflow.
MODEL 09planar 2-link arm · 2 actuators
Gymnasium's planar two-link arm reaching for a target; trains in minutes.
MODEL 10one-legged hopper · 3 actuators
The single-leg hopper from Gymnasium: hop forward without falling.
MODEL 11planar biped · 6 actuators
The classic two-legged planar walker from Gymnasium.
MODEL 12planar runner · 6 actuators
Cannot fall; pure speed and gait efficiency.
MODEL 13quadruped · 8 actuators
Gymnasium's four-legged walker with contact-rich dynamics.
MODEL 143D humanoid · 17 actuators
Gymnasium's full 3D biped — high-dimensional and unstable.
Realistic MuJoCo Menagerie models. The ones marked preview are simulation-only on our CPU servers for now: plan, check and run short tries.
MODEL 15consumer quadruped · 12 actuators · ≈40 cm tall
Unitree's robot dog from MuJoCo Menagerie (position-controlled variant), starting from its standing pose.
Unitree · BSD-3-Clause
MODEL 16Preview · sim onlyindustrial quadruped · 12 actuators · ≈84 cm tall
Spot from MuJoCo Menagerie. A preview robot on kualia: plan tasks, run the checks and short capped tries; full training needs GPU runs.
Boston Dynamics · BSD-3-Clause
MODEL 17Preview · sim onlyfull-size humanoid · 29 actuators · ≈1.3 m
Unitree's 29-joint humanoid from MuJoCo Menagerie. A preview robot on kualia: plan tasks and run short capped tries; a real gait needs GPU-scale training.
Unitree · BSD-3-Clause
MODEL 18Preview · sim onlyresearch arm + gripper · 8 actuators · ≈85 cm reach
The 7-axis research arm with a gripper from MuJoCo Menagerie. A preview robot on kualia with capped runs.
Franka Robotics · Apache-2.0