An open-source harness that lets LLM agents control robots through shared tools, with Codex and Claude Code runtimes and simulator integrations.
pollen-robotics/reachy_mini
Python SDK and app platform for Reachy Mini, an open-source expressive desktop robot for developers and AI builders.
- Stars
- 1,644
- Forks
- 348
- Open issues
- 149
- Last commit
- 2 days ago
What is it?
Reachy Mini is an open-source, expressive robot aimed at hackers and AI builders. This repository is its SDK, which provides Python control of the robot's head pose and movements, along with tooling for running apps on the robot or from a computer.
The SDK supports three setups: the wireless Reachy Mini (Raspberry Pi CM4 with battery and WiFi), the Reachy Mini Lite connected to a PC over USB, and a MuJoCo-based simulation that needs no hardware. Apps can be written in Python for on-robot control loops, or as browser-based JavaScript apps that drive the robot over WebRTC. Community apps are distributed through Hugging Face Spaces and installed from the Reachy Mini Control application.
The repository includes examples, Jupyter tutorial notebooks covering connection, movement, camera and audio, and an AGENTS.md guide for AI coding agents building apps for the robot.
Key features
- 01Python API for controlling head pose and movement with goto_target and related calls.
- 02JavaScript SDK for building browser apps that control the robot over WebRTC.
- 03Supports wireless (Raspberry Pi CM4), USB-connected Lite, and MuJoCo simulation setups.
- 04Hugging Face Spaces app ecosystem with install support via Reachy Mini Control.
- 05Jupyter notebooks and example apps for connection, movement, camera and audio.
- 06Optional uv integration for faster app installation, with pip fallback.
Use cases
- Prototyping robot behaviors in simulation before running them on hardware.
- Building a browser-based web app that lets others control a Reachy Mini robot.
- Connecting an LLM or voice assistant to a physical expressive robot.
- Writing Python control loops for head movement and camera-driven tasks.
Quick start
- 1
Choose your platform guide (Wireless, Lite, or Simulation)
- 2
Follow the Installation guide to set up the SDK
- 3
Run the Quick Look Python example once the robot is awake
Things to know
- Full use requires buying and assembling a physical robot, which takes 2 to 3 hours.
- Hardware design files are licensed under CC BY-SA-NC, which restricts commercial use of the hardware designs.
In collections
Topics
FAQ
- What is reachy_mini?
- Python SDK and app platform for Reachy Mini, an open-source expressive desktop robot for developers and AI builders. Reachy Mini is an open-source, expressive robot aimed at hackers and AI builders. This repository is its SDK, which provides Python control of the robot's head pose and movements, along with tooling for running apps on the robot or from a computer.
- Is reachy_mini free to use? What license does it have?
- reachy_mini is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
- Is reachy_mini still maintained?
- Actively maintained. The last commit to reachy_mini was on Nov 5, 2026.
- What can I use reachy_mini for?
- Prototyping robot behaviors in simulation before running them on hardware. Building a browser-based web app that lets others control a Reachy Mini robot. Connecting an LLM or voice assistant to a physical expressive robot. Writing Python control loops for head movement and camera-driven tasks.
- How do I install reachy_mini?
- Choose your platform guide (Wireless, Lite, or Simulation) Follow the Installation guide to set up the SDK Run the Quick Look Python example once the robot is awake
- Where can I get reachy_mini?
- The source code and installation instructions are on GitHub: https://github.com/pollen-robotics/reachy_mini
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Python framework for building custom FPGA cores and SoCs, with buses, peripherals, CPU integration and simulation support.
Open source SDKs and firmware for building your own Muse gadgets on ESP32 boards and Linux devices like Raspberry Pi.
Open-source KVM over IP hardware and firmware for remotely viewing and controlling a computer's screen, keyboard and mouse.
A simulation platform for robotics and embodied AI combining a multi-physics engine, a photo-realistic renderer, and a cross-platform compiler behind a Python interface.
A lightweight Python driving simulator for reinforcement learning and autonomous driving research, with sensor simulation and Gymnasium-style environments.