Free🇺🇸
Open vision-language-action model from NVIDIA for manipulation tasks on humanoid and other robots, with fine-tuning on your own data.
LeRobot overview and key facts (October 2026)
Open source robotics library from Hugging Face with pretrained policies, a common dataset format and a hardware interface.
Free
What is LeRobot?
LeRobot is Hugging Face's open source library for real-world robotics in PyTorch. It gives developers, labs and small automation teams pretrained policies, a shared dataset format and a common interface to low-cost arms up to humanoids.
The library provides a hardware-agnostic Robot class that separates control logic from the device, with native support for arms such as SO-100 and Koch, mobile platforms such as LeKiwi and humanoids such as Unitree G1. The LeRobotDataset format stores synchronized video and Parquet state data on the Hugging Face Hub, with tools to split, merge and edit episodes.
LeRobot installs with pip and includes training and evaluation of policies, among them SmolVLA, whose base checkpoint had about 72,000 downloads in the last 30 days. The repository had 27,825 GitHub stars and was updated on September 28, 2026.
LeRobot AI features and use cases
- Records and shares robot dataTeleoperation and recording tools capture camera video and robot states in one dataset format. Teams can build training data from demonstrations and reuse thousands of public datasets on the Hub.
- Trains policies from demonstrationsLeRobot trains imitation learning and vision-language-action policies such as SmolVLA on recorded episodes. A small team can teach a low-cost arm a pick task without writing motion code.
- Runs models on different hardwareThe common Robot interface lets the same policy code drive supported arms, mobile bases and humanoids, and custom robots can be added. Teams can switch hardware without rewriting their pipeline.
What sets LeRobot apart
- Supports low-cost arms like SO-100, which makes robot learning affordable for small teams.
- Datasets live on the Hugging Face Hub in one shared format.
- Apache 2.0 license and installation via pip.
Where LeRobot is strongest
Available internationally
LeRobot is an open source project under the Apache 2.0 license and runs on your own hardware in any country.
Questions to ask before you buy LeRobot
LeRobot is a library, not a supported product: your team sets up hardware, trains models and maintains the code. Recordings may show people or workplaces, so decide whether datasets stay local or go to the Hugging Face Hub and check the terms of public datasets and pretrained models you reuse.
LeRobot reviews
No reviews yet. Be the first to share how your team uses it.
Every review shows its verification level. We read each one before it goes live. How we check reviews
Questions about LeRobot
Which robots work with LeRobot?
The README lists SO-100, LeKiwi, Koch, HopeJR, OMX, EarthRover, Reachy2, OpenARM, Unitree G1 and others, plus gamepads and phones for teleoperation. Custom robots can implement the Robot interface.
Is LeRobot free for companies?
Yes. The repository is under the Apache 2.0 license. Pretrained models and datasets on the Hub can carry their own licenses.
Is the project active?
The repository was updated on September 28, 2026, and had 27,825 GitHub stars that day.
LeRobot update historyno updates yet
No changes recorded for LeRobot yet.
Changes the vendor made and confirmed user reports, each checked against the vendor's own pages. We review every tool monthly.
Alternatives to LeRobot
Tools in the same category (Robot foundation models) with the most overlapping markets.
FreeAvailable internationally
Open robotics models π0, π0-FAST and π0.5 from Physical Intelligence with base checkpoints and fine-tuning code.