Skip to content
humanoidrobots.training

Starting viewport

Stanford Student Robotics / Hands-On Robotics · quadruped

Pupper v3

Open-source educational quadruped developed through Stanford courses, with brushless motors, a Raspberry Pi 5 and a reinforcement-learning locomotion policy out of the box.

Frames · Pupper v3

4 items · credit and license on each
  • Pupper v3 studio render, three-quarter front view
    Studio render of the published MJCF, keyframe pose from the model. Geometry only; colours are the model's own materials.Rendered by Hyperspawn from G-Levine/pupper_v3_description (MIT) · Hyperspawn render · source
  • Pupper v3 studio render, side view
    Side view of the same MJCF render, keyframe pose from the model.Rendered by Hyperspawn from G-Levine/pupper_v3_description (MIT) · Hyperspawn render · source
  • 4 s turntable of the MJCF studio render.Rendered by Hyperspawn from G-Levine/pupper_v3_description (MIT) · Hyperspawn render · source
  • Completed Pupper v3 quadruped standing, turquoise printed parts
    From the pupper-v3 documentation.Hands-On Robotics (pupper-v3-documentation) · AGPL-3.0 · source

Provenance

No browser-ready model packaged yet.

The description is public, but nobody has pinned a commit and checked it compiles in MuJoCo WASM. Clone it and drop the folder on the viewport above (MJCF, or URDF with STL/OBJ meshes). Files stay in your tab.

Policies listed for Pupper v3

  • velocity-tracking locomotion from joystick (default walking policy)

    MuJoCo MJX + Brax PPO (pupperv3-mjx), JSON weights run by the neural_controller ROS 2 node · not replayed in the tab

    source

  • three-legged walking

    MuJoCo MJX + Brax PPO (pupperv3-mjx) · not replayed in the tab

    source

Train on GPUs

mjlab ships no stock task for Pupper v3. The training page gives you the pinned model and the steps to write a robot config.

Build the job →