Wendy

Build, deploy, and debug apps on robots, Raspberry Pis, Jetsons, and Linux devices.
Bring your Dockerfile or Compose project. Start with hardware or a simulator.

Get started · Watch the demo · Examples · Discord

Install

On macOS or Linux:

curl -fsSL https://install.wendy.dev/cli.sh | bash
wendy --version

On Windows:

winget install WendyLabs.Wendy --source winget

The Linux app examples below use a macOS or Linux terminal. On Windows, use
WSL2 with the Wendy CLI and Docker with Buildx installed inside WSL2.

Other installation methods and updates

Choose a starting point

Use your own device

wendy tour --hardware

The tour guides you through device setup. Start with a Raspberry Pi, NVIDIA
Jetson, or an existing Linux computer or robot.
Find the setup guide for your hardware.

Try a simulated robot

wendy tour --simulator

The tour creates a simulated Unitree Go2, opens its 3D viewer, and runs a short
walking demo. No physical robot or cloud account is needed. The installer may
start the tour automatically.

You need macOS or Linux, Docker with Buildx, and QEMU. Allow 4 GiB of memory for
the VM and several minutes for the first download and build. Apple Silicon is
the fastest path; x86 Linux uses ARM64 emulation.
See simulator setup
for prerequisites. When finished, use the wendy vm stop command printed by the tour.

Build and deploy an app

With your device set up and Docker running, create a small HTTP app:

wendy init hello-wendy --target wendyos --language node --template simple-api --var APP_ID=hello-wendy --assistant skip --git-init no
cd hello-wendy
wendy run

Select your device. Open the address Wendy prints and check for
{"message":"hello-world"}. Press Ctrl+C, edit src/index.ts, save, then run
wendy run again against the same target. Node.js runs inside the container.
Follow the full walkthrough
for watch mode, local-only Linux targets, and stopping the app when finished.

Why Wendy?

The CLI runs on your computer; Wendy Agent runs on your device,
manages your apps, and returns logs. Install the Agent on an existing Linux system,
or use WendyOS with the Agent included.

• Build and deploy: use Docker or Compose, build for the target, and send changes with wendy run.
• Work with hardware: declare camera, GPU, and device access in wendy.json; inspect ROS 2 nodes and topics.
• Use a coding agent: connect through MCP to deploy changes and inspect the device.

Local development needs no cloud account. Wendy Cloud
adds remote access and fleet management. Compare deployment options.

Build with a coding agent

After your first successful run, you can let a coding agent deploy changes and
inspect the device through Wendy's MCP server:

wendy mcp setup

Follow the coding-agent setup guide
for supported tools. For ChatGPT, see the ChatGPT guide.

Keep building

• Robot apps with ROS 2 in containers, remote topic inspection, and a Foxglove bridge.
• Camera and AI apps with access to cameras, audio, and GPUs.
• Multi-service apps with ROS 2 talkers, listeners, and shared configuration.
• Data collection that records sensor data, app events, and model results around triggers.

Browse the examples or follow a
tutorial to build your next app.
Wendy Cloud adds remote access and fleet management.
Local development does not require a cloud account.