# MuJoCo

Multi-Joint dynamics with Contact. A general purpose physics simulator.

Repository: https://github.com/google-deepmind/mujoco
Canonical: https://ross.abutalabs.com/products/mujoco
Homepage: https://mujoco.org
Language: C++
License: Apache-2.0
License Family: permissive
Topics: robotics, physics, mujoco
Last push: 2026-08-26T22:25:03+00:00
Link (homepage): https://mujoco.org

## Health v2 (maintenance only)
Score: 99/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 100
- inputs: {"age_days": 1832, "days_push": 7, "days_rel": 13, "gap_med": 28.5, "n_releases_24m": 23}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 14680, forks 1698 (observed 2026-08-28T04:11:07.416007+00:00)

## What it is
MuJoCo (Multi-Joint dynamics with Contact) is a general-purpose physics engine for fast and accurate simulation of articulated structures interacting with their environment. It provides a C API, Python bindings, an interactive OpenGL viewer, and a GPU-accelerated variant (MuJoCo Warp), maintained by Google DeepMind.

## Use cases
- simulate robot dynamics with contact for research
- train reinforcement learning agents in physics environments
- simulate biomechanics and human motion
- render physics-based animation for graphics
- run fast GPU-accelerated physics simulation with MuJoCo Warp
- build simulation models from XML with the built-in compiler
- interactively inspect and debug models in the simulate viewer

## When to choose
- you need fast, accurate contact-rich dynamics for robotics or RL research
- you want a permissively licensed (Apache-2.0) physics engine with Python bindings
- you need GPU-accelerated batch simulation via MuJoCo Warp
- you work in biomechanics, animation, or machine learning requiring articulated-body simulation

## When to avoid
- you need soft-body or fluid simulation as a primary feature
- you want a full game-engine with asset pipelines and editing tools out of the box
- you need multiplayer or networked physics out of the box

## Facets
- artifact type: library
- maturity: active
- function: simulation, graphics, machine-learning, sdk
- domain: robotics, simulation, machine-learning, graphics
- platform: cross-platform, cpp, python
- tags: physics-engine, contact-dynamics, articulated-bodies, reinforcement-learning, biomechanics, cuda, unity-plugin, xml-model-compiler, game-development, gpu

## Member repositories
- google-deepmind/mujoco (main) score 99
- google-deepmind/mujoco_menagerie (examples) score 75
- google-deepmind/mujoco_warp (backend) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:07.416007+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:12:25.608235+00:00, confidence not recorded.
  - readme: https://github.com/google-deepmind/mujoco (fetched 2026-08-28T04:11:07.416007+00:00, sha 9f6d0359d30b)
  - homepage: https://mujoco.org (fetched 2026-08-29T08:05:41.808877+00:00, sha dc082c54cacf)
  - registry_pypi: https://pypi.org/pypi/mujoco_warp/json (fetched 2026-08-29T08:05:41.818139+00:00, sha 819d47842f3a)
- Data as of 2026-08-30T08:39:29.467469+00:00.