HuggingFace

Huggingface 下载加速 MCP

javascript

3

GitHub Stars

javascript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "freefish1218-mcp-huggingfetch": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-huggingfetch@latest"
      ],
      "env": {
        "HF_HOME": "~/.cache/huggingface",
        "LOG_LEVEL": "info",
        "HUGGINGFACE_TOKEN": "YOUR_TOKEN_HERE",
        "HUGGINGFETCH_DOWNLOAD_DIR": "~/Downloads/huggingface_models"
      }
    }
  }
}

You can accelerate model file handling with the MCP HuggingFace Fast Download Tool. It enables high-speed, resumable downloads and intelligent retries when pulling assets from HuggingFace, making it ideal for large model files and frequent pulls across Claude Desktop, Claude Code, Cursor, VS Code, and other MCP-enabled clients.

How to use

After you configure the MCP server, you can perform common operations directly in conversations. You can list repository files, explore repository structures, search for specific files, and selectively download models.

Key actions you can perform include:

List Files: View and filter repository files to locate what you need.
Explore Repository: Understand the directory structure and get a high-level file tree.
Search Files: Find specific files by name or pattern within a repository.
Download Models: Choose which files to download and specify target directories. You can constrain downloads by file types or patterns.

How to install

Prerequisites: Node.js 18+ and npm or yarn.

Clone the MCP HuggingFace Fast Download Tool repository.
Change into the project directory.
Install dependencies.

git clone https://github.com/freefish1218/mcp-huggingfetch.git
cd mcp-huggingfetch
npm install

Configuration examples for MCP clients

The following configuration snippets show how to enable the HuggingFace download tool in different MCP-supported clients. Each snippet sets the access token through an environment variable.

{
  "mcpServers": {
    "huggingfetch": {
      "command": "npx",
      "args": ["-y", "mcp-huggingfetch@latest"],
      "env": {
        "HUGGINGFACE_TOKEN": "your_token_here"
      }
    }
  }
}

Environment and credentials

To access private models or private repositories, obtain a HuggingFace access token and configure it in your MCP client as shown above.

Steps to get a token:

Go to HuggingFace Settings
Create a new Access Token
Copy the token into HUGGINGFACE_TOKEN in your MCP server configuration

Usage examples

List JSON files in the 2Noise/ChatTTS repository

Explore the directory structure of microsoft/DialoGPT-medium

Search for config files in openai/whisper-large-v3

Please download the ChatTTS model to ./models directory

Download microsoft/DialoGPT-medium model, only .bin files

Download openai/whisper-large-v3 model, exclude test files

Available tools

list_huggingface_files

List and filter repository files with pattern matching and sorting.

explore_huggingface_repo

Generate a hierarchical view of a repository's file tree and structure.

search_huggingface_files

Search for files by name or pattern within a repository.

download_huggingface_model

Download specific files from a repository to a designated local directory with include/exclude patterns.