LeRobot documentation
LeRobot
LeRobot
State-of-the-art machine learning for real-world robotics
🤗 LeRobot provides a hardware-agnostic, Python-native interface for controlling real robots - from affordable arms like the SO-ARM101 to full humanoids. Plus the tools to record, store, and share the datasets they generate. Every dataset uses the standardized LeRobotDataset format (synchronized video + action/state data) and can be streamed directly from the Hugging Face Hub.
🤗 On top of that data, LeRobot implements state-of-the-art policies - from lightweight imitation-learning models like ACT to large vision-language-action models like π₀ and SmolVLA - all trainable, shareable, and deployable with the same handful of CLI commands.
The goal: lower the barrier to entry for robotics, so that everyone can contribute to, and benefit from, shared datasets and pretrained models.
How It Works
Teleoperate → Record → Train → Deploy
- Teleoperate - control the robot yourself (with a leader arm, keyboard, or phone) so it can learn from your movements.
- Record - each demonstration is saved as a dataset: synchronized camera video plus the actions you took.
- Train - a policy (the neural network that will control the robot) learns to imitate your demonstrations.
- Deploy - run the trained policy on the robot and watch it complete the task on its own.
Get Started
New here? Install LeRobot, then pick your path:
LeRobot supports a wide range of arms and mobile robots. Popular picks:
- SO-101 - our flagship, low-cost arm
- LeKiwi - a mobile base with an arm on top
- Koch v1.1 - a long-time community favorite
- or find yours under Robots in the sidebar
Once it's assembled and calibrated, record a dataset and train your first policy with the imitation learning tutorial - or skip the CLI entirely with LeLab, a browser GUI for the same workflow.
You can still train and evaluate policies without owning a robot:
- train on an existing LeRobot dataset from the Hub
- evaluate in simulation, against benchmarks like LIBERO or Meta-World
- try the free Colab notebooks - nothing to install
Start with the Contributing guide, then add a new policy or bring your own hardware.
Explore the Docs
Every LeRobot CLI command, copy-paste ready.
Which policy fits your GPU, and how long training takes.
Load, stream, and visualize robot datasets from the Hub.
A browser GUI for calibrating, recording, and training - no CLI required.
Start with ACT, our recommended first policy - or browse SmolVLA, π₀, and more in the sidebar.
Train and evaluate in simulated environments before touching real hardware.
Common Problems
Running into issues? A few of the most frequent ones:
- Blurry or unusable camera footage - lighting matters more than resolution. See the Cameras guide.
- Build or install errors (
cmake,ffmpeg, CUDA) - see the Troubleshooting section of the Installation guide. - Not sure which policy fits your GPU - check the Compute & Hardware Guide.
- Still stuck? Ask on Discord - the community (and the LeRobot team) is there to help.