Linux for Robots,
without the pain.
WendyOS is the Apache 2.0 open source operating system and developer tools for NVIDIA Jetson and Raspberry Pi, with Wendy for headless Mac in progress. Ship physical AI apps faster than ever.
➜ ~ wendy discover
✔︎ Discovering all Wendy devices [0.5s]
✔ Success
Found 1 Wendy device
Humble Pepper (version: 2025.12.17)
USB 3.2: VID: 0x1D6B, PID: 0x0104
Ethernet: en30, 2.5 Gbps
LAN: wendyos-humble-pepper.local
➜ ~ wendy run
✔︎ Searching for WendyOS devices [0.2s]
✔︎ Builder ready [0.1s]
✔︎ Container built and uploaded successfully! [2.3s]
✔︎ App ready to start [0.1s]
✔ Success
Started app
USB-C Development
No WiFi? No Problem.
Just plug in your device via USB-C. WendyOS handles the networking, letting you deploy and debug as if the device was running locally on your machine.
Modern Toolchain
Swift, Python, Rust & C++
Write in your favorite language. WendyOS treats Swift as a first-class citizen for safety and performance, while fully supporting Python, Rust, and C++ containers.
Remote Access
Debug from Anywhere
Access logs, remote shells, and process status from anywhere in the world. Perfect for managing fleets of robots or distributed sensor networks.
OTA Updates
Fleet Management Built-in
Push application updates or full OS upgrades to one device or ten thousand. Atomic updates ensure your devices never get bricked.
Hardware Accelerated
Jetson, Raspberry Pi & Mac
Optimized for edge hardware today, with Wendy for headless Mac bringing the same deploy and debug workflow to Mac mini and Mac Studio targets.
Physical AI Ready
Sensors & Actuators
Built for the physical world. Simple APIs for GPIO, I2C, SPI, and Camera access make controlling motors and reading sensors straightforward.
A modern IDE for
embedded development
Stop wrestling with SSH keys, cross-compilation toolchains, and flaky connections. The Wendy VSCode extension treats your remote device like a local simulator. Available on the VS Code Marketplace and Open VSX Registry (for Cursor, Windsurf, etc).
One-Click Deploy
Build and push containers instantly.
Remote Debugging
Full LLDB & Python debugging support.
Polyglot
Swift, Python, Rust, C++, TypeScript.
Hardware Discovery
Auto-detect devices via USB or LAN.
2
3
4
5
6
7
8
from ultralytics import YOLO
# Load model onto GPU
model = YOLO("yolov8n.pt")
# Start inference stream
results = model.predict(source=0, stream=True)
for r in results:
process_frame(r.orig_img)
✔︎ Container built and uploaded successfully!
ℹ︎ Debugger attached on port 5678
Loaded YOLOv8n model...
Inference running at 30 FPS
Built for the best
AI Hardware
Wendy targets devices with strong local compute, from NVIDIA Jetson and Raspberry Pi today to headless Apple Silicon Macs as Wendy for Mac lands.
NVIDIA Jetson
Orin Nano, AGX Orin, Thor
- Unified Memory Architecture
- MLX & PyTorch CUDA Support
- Hardware Video Encoding
- 275-2000 TOPS AI Performance
Raspberry Pi
Pi 5, Pi 4, Pi 3
- Low Power Consumption
- Broad GPIO Ecosystem
- Hardware PWM & SPI
- Affordable Entry Point
Headless Mac
Mac mini, Mac Studio
- Wendy Agent Workflow
- Apple Silicon Performance
- Remote Deploy & Debug
- In Progress for Mac Targets
Ready to Build Your Next Physical Intelligence Startup?
Whether you are upgrading from Orin or starting fresh on Thor, our engineers are ready to help you ship your next physical intelligence app.
Get Wendy