# unitreerobotics/unitree_rl_gym

Repository: https://github.com/unitreerobotics/unitree_rl_gym
Canonical: https://ross.abutalabs.com/products/unitree_rl_gym
Language: Python
License: BSD-3-Clause
License Family: permissive
Last push: 2025-07-25T10:20:22+00:00

## Health v2 (maintenance only)
Score: 42/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 33, release rhythm 35, longevity 75
- inputs: {"age_days": 1057, "days_push": 404, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3516, forks 580 (observed 2026-08-28T04:08:07.846353+00:00)

## What it is
A reinforcement learning gym environment for training locomotion policies on Unitree robots (Go2, H1, H1_2, G1) using Isaac Gym and Mujoco. It provides a train → play → sim2sim → sim2real workflow for deploying learned motion control to physical robots.

## Use cases
- train locomotion policies for Unitree quadruped and humanoid robots
- reinforcement learning for legged robot motion control
- sim-to-real transfer of robot control policies
- train walking policies in Isaac Gym and deploy to hardware
- validate RL policies across simulators with sim2sim
- learn reinforcement learning for robotics with real robots

## When to choose
- you own or target Unitree Go2, H1, or G1 robots
- you want an end-to-end RL-to-real-robot pipeline for legged locomotion
- you need GPU-parallel training environments for robot policies

## When to avoid
- you need RL for non-Unitree robots or manipulators
- you only want simulation without real-robot deployment
- you lack an NVIDIA GPU required by Isaac Gym

## Facets
- artifact type: library
- maturity: active
- function: reinforcement-learning, simulation, robotics, machine-learning, llm-training
- domain: robotics, reinforcement-learning, simulation, machine-learning
- platform: python, cross-platform
- tags: legged-robots, sim2real, isaac-gym, mujoco, motion-control, unitree, gym-environment, quadruped, humanoid, linux, gpu

## Member repositories
- unitreerobotics/unitree_rl_gym (main) score 42

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:07.846353+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 2026-08-29T18:35:35.389254+00:00, confidence not recorded.
  - readme: https://github.com/unitreerobotics/unitree_rl_gym (fetched 2026-08-28T04:08:07.846353+00:00, sha a6d8afd9d610)
- Data as of 2026-08-30T08:39:29.467469+00:00.