mjlab
- Software
- Humanoids and robot learning
- Framework
A framework for robot learning that pairs Isaac Lab's manager-based API with GPU-accelerated MuJoCo Warp, for designing simulation environments and training policies.
Source on GitHub: github.com/mujocolab/mjlab (opens in a new tab). Facts read at commit 8bb2bc9 (opens in a new tab).
Licenses
- Code
- Apache-2.0 (opens in a new tab)
Read by Robotheca from LICENSE (opens in a new tab) (Apache-2.0) at commit 8bb2bc9 (opens in a new tab), on 23 Oct 2026.
- Commercial use
- Yes
What it does: Robotheca's summary
- Provides composable building blocks for designing robot learning environments with minimal dependencies.
- Gives direct access to native MuJoCo data structures.
- Includes a velocity-tracking task that trains a Unitree G1 humanoid to follow velocity commands on flat terrain.
- Includes a motion-imitation task that trains a humanoid to mimic reference motions.
- Supports multi-GPU training via a gpu-ids option.
- Offers built-in zero and random agents to sanity-check an MDP before training.
Written by Robotheca from the project's README (opens in a new tab) at commit 45b1e26.
Getting started
- Difficulty
- Advanced
- Hardware it needs, as the project states it
- Requires an NVIDIA GPU for training; macOS is supported for evaluation only.
- How to start
- The project's website (opens in a new tab)
Maintenance
- Status
- Active: a commit in the last 90 days
- Newest commit
- 18 Oct 2026 on main
- Latest release
- mjlab v1.6.0 (9 Aug 2026)
Checked on GitHub on 30 Oct 2026.
Links
Robots that can run it
No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.
Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.