mujoco-python

IsaacLab to Mujoco GO2 deploy, IsaacLab to Real world GO2 deploy

Reinforcement learning policies trained in NVIDIA Isaac Lab & MuJoCo Warp for robots: asimov-1 (humanoid, PPO+AMP omni-gait), v1x1 (quadruped, PPO omni-gait), BDX-R (biped, omni-gait) and Microduck (biped, omni-gait & fall recovery) with live ROS2 teleop bridges for real-time keyboard control and other sim-to-real pipelines..

Utils for editing mujoco xml

Rust client for the Franka Control Interface: one binary for the FR3 and the Panda, no C++, the 1 kHz loop as a library and low rate interface. Python bindings included.

Utility helpers for MuJoCo simulations in Python.

Simulation-based design and hybrid control of a two-wheeled self-balancing jumping robot in MuJoCo using P-V state feedback, phase-dependent leg control and contact-based event detection.

Grasp policies for any robotic hand

Simple renderer for use with MuJoCo (2.2.x) Python Bindings, on M1 Mac.

Robotic assembly cell simulation of a Keychron K2 mechanical keyboard implemented in MuJoCo

Motion control and balance of a two-wheeled inverted pendulum Robot using Reinforcement Learning

Sparse, Residual World Models for Object-Centric Manipulation — code, paper, and browser demo

Simple Flow Matching Policy for Single-Arm Manipulator with Obstacle Avoidance

A stub file for mujoco's python interface (wip)

Wrappers for environments from Mujoco Gym and dm_control that make them deterministic

Demo showcasing C++ code linking against MuJoCo, consuming data from Python without copies

Gestured controlled humanoid arm, precise movement controlled via hand pose detection(mediapipe HandLandmarker)

Teaching a simulated quadruped robot how to run and analyse training and learned behavior

A complete MJCF lifecycle and trial orchestration suite for MuJoCo, powered by Pydantic v2.

Compare rule-based, TSK fuzzy, and learned behaviour-cloning controllers using deterministic simulation, noisy sensor models, dataset collection, and automated evaluation.