Open software that robots run on, with the license read from each project's license file, whether it is still maintained, and the robots on Robotheca that use it. A reference, not a download site: every entry links to its source.
Page 1 of 2 · newest commit first
Open-source SDK for Reachy Mini, an expressive desktop robot kit aimed at hackers and AI builders, for scripting its head, camera, audio and behaviors.
LeRobot is a PyTorch library of models, datasets, and tools for real-world robotics, aiming to lower the barrier to entry for shared datasets and pretrained models.
ROS 2 packages that embed llama.cpp so GGUF language, vision-language, audio and embedding models can run inside ROS 2 projects.
An open-source modular Python library for reinforcement learning, implemented in PyTorch, JAX and NVIDIA Warp, focused on modularity, readability and transparency.
DimOS is a Python-based agent-native operating system for physical robots, built to run on humanoids, quadrupeds, and drones without requiring ROS.
Unitree's second-generation C++ software development kit for controlling Unitree robots. It provides the library and examples needed to build applications that talk to the robots.
A framework for training and deploying reinforcement learning control policies on humanoid robots, plus converting human motion capture data into robot motions.
Stable-Baselines3 is a set of reliable PyTorch implementations of reinforcement learning algorithms, meant as baselines for research and for building new projects.
Crocoddyl is an optimal control library for robots under contact sequences, using differential dynamic programming solvers to compute optimal trajectories and feedback gains.
A lightweight, GPU-accelerated Python library for robot learning research, offering PPO and student-teacher distillation with minimal, easy-to-extend code.
RAI is an agent framework for building and deploying embodied AI features on robots, with multi-agent, human-robot interaction and multi-modal support.
A Python library of reinforcement learning robot environments built on the Gymnasium API and the MuJoCo physics engine, covering manipulation, dexterous hands, multi-agent and maze tasks.
An open benchmark of simulated robotic manipulation tasks for developing and evaluating multi-task and meta reinforcement learning algorithms.
TSID is a C++ library for optimization-based inverse-dynamics control of robots, built on the Pinocchio rigid multi-body dynamics library, with Python bindings.
A simulation benchmark for evaluating robot manipulation policies against real-world setups, built on SAPIEN and ManiSkill2 with Bridge dataset environments in ManiSkill3.
A modular AI runtime for building multimodal agents that run on digital platforms and physical robots, from humanoids and quadrupeds to educational and simulated robots.
ROS 2 packages for the ROBOTIS AI Worker robot, providing interfaces and tools to control it, integrate its sensors, and build physical AI applications.
An MCP server that lets large language models such as Claude, GPT, and Gemini talk two ways with ROS robots, without changing the robot's existing source code.
A GPU-accelerated framework for training physically simulated digital humans and humanoid robots, aimed at animation, robotics, and reinforcement learning research.
GELLO is a low-cost, intuitive teleoperation framework that lets a person control robot manipulator arms using a scaled-down leader arm. It covers several arms, including the I2RT YAM.
An open-source Arduino and Raspberry Pi framework for programming quadruped robots, covering gait, servo control and IMU integration; it targets the legacy NyBoard platform.
A MuJoCo-based simulator for Unitree robots that lets controllers built with Unitree SDK2, unitree_ros2 and unitree_sdk2_python run in simulation before real hardware.
A framework for robot learning from demonstration, offering datasets of robot manipulation demonstrations and offline learning algorithms to train policies from them.
Stretch AI is a software toolkit for building intelligent behaviors on the Stretch 3 mobile manipulator, covering grasping, mapping, navigation, LLM agents, speech, and embodied question answering.
Projects 1–24 of 36