> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/ai-workbench/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/ai-workbench/_mcp/server.

> Instructions for installing NVIDIA AI Workbench on a Jetson device.

# Install AI Workbench on Jetson

<Note title="Downloading or installing AI Workbench implies consent to the NVIDIA AI Product Agreement.">

By downloading or installing the NVIDIA AI Workbench software, you agree to the terms of the [NVIDIA AI Product Agreement](https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/) (EULA).
If you do not agree to the terms of the EULA, you are not authorized to download, install, or use NVIDIA AI Workbench.

</Note>

## Overview

**You can install AI Workbench on a Jetson device using the CLI.**

You cannot install the Desktop App on a Jetson device.
After installation, you add the Jetson as a remote location in the Desktop App on your local machine.

**Jetson requires Docker as the container runtime and has a known compatibility issue with JetPack.**

The installation process includes a step to resolve the Docker compatibility issue.
This involves creating a configuration file that adjusts how Docker handles networking on the device.

**You need SSH access to the Jetson device before you can install AI Workbench.**

The installation process requires you to run commands directly on the device.
You also need SSH access later to connect the Jetson as a remote location in the Desktop App.

## Key Concepts

**JetPack:**

The NVIDIA SDK for Jetson devices.
Includes a known compatibility issue with newer Docker versions that requires a configuration workaround.

**AI Workbench Application Folder:**

The destination folder on the Jetson for binaries and program files.
The name is fixed as *.nvwb*, and the default location is in the home directory, *~/.nvwb*.

## Prerequisites

**Your Jetson device must meet the following system requirements.**

| **Resource** | **Size** | **Description** |
|---|---|---|
| Disk space | `500 MB` | Minimum required for AI Workbench application files. |
| Disk space | `40 GB` | Minimum required for development containers. |
| RAM | `16 GB` | Minimum required to run AI Workbench and development containers. |

**You also need the following.**

- Your username on the Jetson device
- Your password for that user
- A local machine with the AI Workbench Desktop App installed

## Configure SSH Access

**Step One: Find the Jetson device's IP address.**

1. Open a terminal on the Jetson device
2. Run `ifconfig`
3. Find the interface that includes `flags=4163<UP,BROADCAST,RUNNING,MULTICAST>`
4. Note the `inet` address (e.g., `192.168.40.125`) for that interface

**Step Two: Install and configure SSH on the Jetson.**

1. Install *openssh-server* on the device

   ```bash
   sudo apt update && sudo apt install openssh-server
   ```

2. Open the required ports for AI Workbench

   ```bash
   sudo ufw enable && sudo ufw allow 10000 && sudo ufw allow 10001
   ```

**Step Three: Set up SSH key access from your local machine.**

1. Check for existing SSH keys on your local machine

   ```bash
   ls -la ~/.ssh/
   ```

2. If no keys exist, create them

   ```bash
   ssh-keygen -t rsa
   ```

3. Add your private key to the SSH agent

   ```bash
   eval "$(ssh-agent -s)"
   ssh-add ~/.ssh/id_rsa
   ```

4. Copy your public key to the Jetson device

   ```bash
   ssh-copy-id <jetson-username>@<jetson-ip-address>
   ```

**Step Four: Test the SSH connection.**

1. Connect to the Jetson from your local machine

   ```bash
   ssh <jetson-username>@<jetson-ip-address>
   ```

2. Verify you can connect without being prompted for a password

**Success: You have SSH access to the Jetson device from your local machine.**

## Install AI Workbench

**Step One: SSH into the Jetson device.**

```bash
ssh <jetson-username>@<jetson-ip-address>
```

**Step Two: Copy and paste the installation command in the terminal.**

```bash
mkdir -p $HOME/.nvwb/bin && \
curl -L https://workbench.download.nvidia.com/stable/workbench-cli/$(curl -L -s https://workbench.download.nvidia.com/stable/workbench-cli/LATEST)/nvwb-cli-Linux-arm64 --output $HOME/.nvwb/bin/nvwb-cli && \
chmod +x $HOME/.nvwb/bin/nvwb-cli && \
sudo -E $HOME/.nvwb/bin/nvwb-cli install
```

**Step Three: Follow the prompts in the TUI and select Docker as the container runtime.**

**Success: AI Workbench is installed on the Jetson device.**

## Fix Docker Compatibility Issue

**Jetson has a known compatibility issue with newer Docker versions that requires a configuration file.**

**Step One: Create the Docker override directory and configuration file.**

```bash
sudo mkdir -p /etc/systemd/system/docker.service.d
sudo nano /etc/systemd/system/docker.service.d/jetson-docker.conf
```

**Step Two: Add the following content to the file.**

```ini
[Service]
Environment="DOCKER_INSECURE_NO_IPTABLES_RAW=1"
```

**Step Three: Reload and restart Docker.**

```bash
sudo systemctl daemon-reload
sudo systemctl restart docker.service
```

**Success: Docker is configured to work with JetPack on the Jetson device.**

## Add Jetson as a Remote Location

**After installing AI Workbench on the Jetson, add it as a remote location in the Desktop App.**

You can add it using the Desktop App or CLI on your local machine.
Follow the steps in the [Remote Locations Guide](/ai-workbench/reference/remote-locations).

## Related Topics

- [Manually Install on Ubuntu (Remote)](/ai-workbench/install/remote-install) - Manual remote install on Ubuntu
- [Remote Locations](/ai-workbench/reference/remote-locations) - Adding remote locations