HybridRobotics/berkeley-humanoid-lite

Health v2 · maintenance only

Adoption not part of the score

1980 stars · 317 forks observed · 2026-10-17

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Berkeley Humanoid Lite is the open-source codebase for a sub-$5,000 3D-printed humanoid robot platform from UC Berkeley. It includes Isaac Lab environments for reinforcement learning policy training, sim2sim validation, real-robot deployment, motion capture, and teleoperated manipulation.

Use cases

• train humanoid locomotion policies with reinforcement learning
• deploy RL policies from simulation to real humanoid hardware
• validate policies across simulators before hardware deployment
• teleoperate a humanoid robot for manipulation tasks
• export robot descriptions (URDF, MJCF, USD) from Onshape CAD
• build a low-cost open-source humanoid robot for research

When to choose

• you want an affordable, open-source humanoid robotics research platform
• you need a full sim-to-real pipeline for legged/humanoid locomotion
• you want to customize or extend a 3D-printed humanoid design
• you are doing RL-based locomotion research with Isaac Lab

When to avoid

• you need a high-payload industrial humanoid or commercial-grade hardware
• you only want a simulator without real robot deployment
• you are not prepared to fabricate 3D-printed parts and assemble hardware
• you need closed-source turnkey humanoid products with vendor support

Facets

library · maturity active

reinforcement-learning simulation robotics machine-learning sdk robotics machine-learning reinforcement-learning simulation hardware python cross-platform humanoid-robot sim2real isaac-lab 3d-printed-robot locomotion teleoperation motion-capture open-source-hardware linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="HybridRobotics/berkeley-humanoid-lite")