18.10.25 - New Release - V0.1.10

View on PyPi

> pip install

eigen-robotics

Control any robot, connect any sensor, instantly switch from sim to real. All in python.

Thank you! Your submission has been received, confirmation will arrive at your email soon.
Oops! Something went wrong while submitting the form.
BETA DOWNLOADS
1089
NODES ONLINE
1074
DEPLOYMENT TIME
<10s
Currently active in robotics development at multiple labs internationally
OE logo2020INC logoThe Paak logo2020INC logo

Instant Sim to Real

Run the same code across MuJoCo, Genesis, and PyBullet — no rewrites, just toggle one config value

Node Based, Modular

Orchestrate distributed robotics compute through a modular, node-based graph.

Built for ML Workflows

Seamlessly integrates data, training, and control in one continuous pipeline.

1.0 - Why Eigen?

Robotics complexity is exploding.

Robotics is growing more complex by the day — with new sensors, models, and control stacks emerging faster than teams can integrate them. Without the right infrastructure, development slows, costs surge, and innovation stalls. To break through the Robotics Complexity Wall, teams need an engine built for speed — one that unifies simulation, learning, and deployment into a single streamlined workflow. Eigen is that engine — designed to help modern teams move faster, outpace legacy systems, and deliver real-world impact.

Read our Whitepaper

eigen-robotics

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Jupyter Notebook & Gymnasium

Bring your favorite ML tools — it just works.

Cross Embodiment Support

Connect to any robot and sensor

Multi-OS Compatibility

Run anywhere — macOS or Ubuntu.

GUI Toolkit

Build, debug, and deploy robotics pipelines through an interactive UI.

Sim-to-Sim & Sim-to-Real

Use one codebase for all testing and deployment

ROS Support

Continue using your existing ROS tools

2.0 - How do I use Eigen?

Get Started Now:

github.com/eigen-robotics

FAQs

> Which Robots and Sensors does Eigen support?

Eigen works with any robot or sensor — our architecture is fully modular. Out of the box, we support Franka, Unitree (Go2 & A1), Interbotix Viper/X-Series arms, RealSense, and standard USB/IP cameras.

You can also plug in your own hardware with a simple Python driver — no ROS or C++ required. If it streams joint positions or sensor data, Eigen can control it.

> Can it integrate with reinforcement learning, diffusion policies etc.?

Yes — Eigen is built for ML workflows from day one. You can collect data, train, deploy, and close the sim-to-real loop inside one framework.

> What makes Eigen different from ROS/LeRobot etc.?

ROS is powerful but heavy and painful to maintain. Newer frameworks are lightweight but locked to one robot or one simulator. Eigen combines the flexibility of ROS with the developer experience of PyTorch — plug-and-play, any robot, any sensor, any simulator.

> How long is the Waitlist?

We’re onboarding teams in batches every few weeks — and we genuinely want to get you in soon. The fastest way to move up the queue is to sign up with your full lab or startup (we prioritise active teams).

Even if you're not ready yet, join the email list — you'll get early releases, private demos, and priority invites as slots open.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.