---
title: "Register a Hugging Face account and get an access token"
description: "Create a Hugging Face account, accept a gated model's license, generate an access token, and save it in ARX so AI Model Center can download the model"
canonical_url: "https://docs.ardge.com/en/knowledge/arx-os/huggingface-token"
---
# Register a Hugging Face account and get an access token

> Create a Hugging Face account, accept a gated model's license, generate an access token, and save it in ARX so AI Model Center can download the model

## Overview

The open-source models in AI Model Center come from Hugging Face. Some of them require you to accept a license before you can download them. These are called gated models, and downloading one needs a Hugging Face account and an access token.

This article covers creating a Hugging Face account, accepting a model's license, generating an access token, and saving that token in ARX.

## What you'll learn

- Create a Hugging Face account
- Accept a gated model's license on its model page
- Generate an access token with read permission for gated models
- Save the token in ARX so AI Model Center can use it to download models

**Estimated time: about 10 minutes**

## Applies to

- ARX OS 1.0 and later

## When do you need a token?

When AI Model Center downloads a model that requires a license, the prompt you see depends on what is already in place:

<table>
<thead>
  <tr>
    <th>
      Situation
    </th>
    
    <th>
      What the system tells you
    </th>
  </tr>
</thead>

<tbody>
  <tr>
    <td>
      No token saved in ARX yet
    </td>
    
    <td>
      Asks you to supply a Hugging Face token
    </td>
  </tr>
  
  <tr>
    <td>
      Token saved, but the model's license not yet accepted
    </td>
    
    <td>
      Asks you to accept the model's license on Hugging Face
    </td>
  </tr>
</tbody>
</table>

In other words, downloading a gated model needs both: a valid token saved in ARX, and the model's license accepted on Hugging Face.

## Step 1: Create a Hugging Face account

1. Go to [huggingface.co](https://huggingface.co) and click **Sign Up**
2. Enter your email address and a password to register
3. Open the verification email and complete verification

Already have an account? Skip this step and sign in.

## Step 2: Accept the model's license

1. Search for the model on Hugging Face and open its model page
2. If the model is gated, the license and an acceptance button appear at the top of the page. The button varies by model — **Agree and access repository** and **Request access** are the common ones
3. Read the license, then click the button

<warning>

**Licenses are accepted one model at a time:**

Every gated model needs its own acceptance on its own model page — accepting one does not cover the others. Some models are moderated, so you can only download them after the author approves your request.

</warning>

## Step 3: Generate an access token

1. Sign in to Hugging Face and go to **Settings › Access Tokens**, or open [huggingface.co/settings/tokens](https://huggingface.co/settings/tokens) directly
2. Click **New token**
3. Name the token and select the permission **Fine-grained** / **Read access to contents of all public gated repos you can access**
4. Copy the token once it is generated

<warning>

**Copy and store the token immediately:**

Save the token as soon as it is created. If you don't, go to the **Access Tokens** page and click **Manage** to refresh it — that generates a new token and invalidates the old one.

</warning>

<tip>

**Create a separate token for each purpose:**

This is Hugging Face's own recommendation. Give ARX a token of its own so that revoking it later doesn't affect anything else you use.

</tip>

## Step 4: Save the token in ARX

1. Sign in to the ARX admin interface and open **AI Model Center**
2. Go to the **Settings** page
3. In the **Open-source Model Platform Token** section, click the add icon on the **Hugging Face** entry
4. Paste the Hugging Face token you created in Step 3
5. Click save

## Related articles

- [Deploy your fine-tuned model on an ARX device](https://docs.ardge.com/en/knowledge/arx-os/deploy-fine-tuned-model)


## Sitemap

See the full [sitemap](https://docs.ardge.com/sitemap.md) for all pages.