> 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.

# Connect NVIDIA Integrations

## Overview

**AI Workbench provides integrations for NVIDIA materials and services to enhance the user experience.**

The integrations streamline authentication needed to access those materials and services.

The integrations are not necessary, and you can have a complete user experience without them.

**Currently there are three NVIDIA integrations you can set: Brev, NGC and Endpoints.**

Configuring NVIDIA Brev enables launching AI Workbench in GPU and CPU instances in the cloud.
Configuring NVIDIA NGC enables pulling containers and models from private registries.
Configuring NVIDIA Endpoints enables using NVIDIA's free inference endpoints.

**You need an NVIDIA account to configure the integrations.**

You can make the account as part of doing the integration in AI Workbench or use an existing account.
The account creation steps are the same no matter which integration you use.

## Key Concepts

**NVIDIA Account**

Single sign-on required to handle authentication or API key creation for the NVIDIA integrations.
Gives you access to Brev, NGC and the Endpoints.

**NVIDIA Brev**

[NVIDIA's platform](https://developer.nvidia.com/brev) that connects you to cloud providers to find and launch the GPU you want to use.
Integrates with AI Workbench to spin up and manage cloud-based AI Workbench locations.

This integration uses an Oauth flow to authenticate your Desktop App to the Brev platform.

**NVIDIA NGC**

[NVIDIA's enterprise platform](https://ngc.nvidia.com) for AI and digital twin workflows with container registry, model catalog, and build services.
Integrates with AI Workbench so you can pull containers from private registries.

This integration uses a scoped API key that you create in the NGC console.

**NVIDIA Endpoints**

[NVIDIA's platform](https://build.nvidia.com) for finding and testing open models on optimized hardware.
Integrates with AI Workbench so you can use free endpoints in the Build Assistant.

This integration uses an API key that you create in the Build console.

Creating a key requires phone number validation.

## Create an NVIDIA Account

If you have an existing account, you should skip this and do the individual integrations with your existing credentials.

You can also do the account creation as part of the integration set up.

**If you are creating an NVIDIA account for the first time, there are a few steps.**

Any NVIDIA platform requires you to do a few things to get started.

**Step One: Initate account creation.**

1. Go [here](https://account.nvidia.com/)
2. Enter your email and create a password
3. Confirm account creation

**Step Two: Verify your email.**

1. Check your email for a confirmation code
2. Enter the code to verify account creation

**Step Three: When prompted, create your NVIDIA Cloud Account.**

1. Give your cloud account a name
2. Accept any required terms

**Step Four: Log in again to finalize account creation.**

1. Log in with your new NVIDIA account credentials

## Connect the NVIDIA Brev Integration

**Step One: Open the AI Workbench Settings window and select the Brev card.**

1. Select **Gear Icon (Top Right) > Settings**
2. Select **Settings > Integrations > Brev Card > Connect**

**Step Two: Enter your email and complete the authentication flow.**

1. Enter your email address
2. Follow the prompts in your browser
3. Wait for the integration to confirm

**Success: The Brev card in Settings shows a connected status.**

## Connect the NVIDIA NGC Integration

**Step One: Open the AI Workbench Settings window and select the NGC card.**

1. Select **Gear Icon (Top Right) > Settings**
2. Select **Settings > Integrations > NGC Card > Connect**

**Step Two: Follow the instructions in the modal.**

1. Go to the [NGC API-Key page](https://org.ngc.nvidia.com/setup/api-keys)

2. Select **Generate Personal Key**

3. Enter a name for your key

4. Set an expiration date (default is 12 months)

5. Select the following services:

   - **NGC Catalog** — Access to the NGC Catalog
   - **NVIDIA Private Registry** — Access to the NGC Container Registry

6. Select **Generate Personal Key**

7. Copy the generated key

**Step Three: Add the Personal Key to AI Workbench.**

1. Paste your key into the API Key box in the modal
2. Select **Submit**

**Success: The NGC card in Settings shows a connected status.**

## Connect the NVIDIA Endpoints Integration

**Step One: Open the AI Workbench Settings window and select the Endpoints card.**

1. Select **Gear Icon (Top Right) > Settings**
2. Select **Settings > Integrations > Endpoints Card > Connect**

**Step Two: Follow the instructions in the modal.**

1. Go to the [Build API-Key page](https://build.nvidia.com/settings/api-keys)
2. Select **Generate API Key**
3. Verify your phone number if requested
4. Enter a name for your key
5. Set an expiration date (default is 12 months)
6. Select **Generate Key**
7. Select **Copy API Key**

**Step Three: Add the API Key to AI Workbench.**

1. Paste your key into the API Key box in the modal
2. Select **Submit**