Hugging Face MCP

Hugging Face MCP is an MCP server on top of the Hugging Face Hub API.

Tools

• search_models: Search models on the Hugging Face Hub. Equivalent to hf_api.list_models.
• get_model_info: Get structured metadata about a model on the Hugging Face Hub. Equivalent to hf_api.model_info.
• get_model_card: Uses hf_api.hf_hub_download to get the full model card (README.md) of a model on the Hugging Face Hub.
• update_metadata: Opens a pull request using hf_api.create_commit to update the metadata of a model on the Hugging Face Hub. Currently only supports updating library_name and pipeline_tag.

Usage

This MCP is tested using Cursor.

Run the Server

1. Clone the repository.
2. Create a virtual environment and install the dependencies.

uv venv
.venv\Scripts\activate # Windows
source .venv/bin/activate # Linux/MacOS
uv pip install fastmcp huggingface_hub

3. Run the MCP server.

python src/server.py

Use the Server in Cursor

1. Press Ctrl+Shift+J -> MCP -> Add new global MCP server to open mcp.json.
2. Add the following configuration:

{
    "mcpServers": {
        "huggingface": {
            "url": "http://127.0.0.1:8000/mcp"
        }
    }
}

3. The server should now be available in the MCP server list, with the tools search_models, get_model_info, get_model_card, and update_metadata.
4. Use Cursor Chat with a frontier model like claude-3.7-sonnet and Agent mode to use the tools.

Examples

• List the top 50 trending models on Hugging Face
• Get the model card for the meta-llama/Meta-Llama-3-8B-Instruct model
• Find the top trending models with gguf in the name and write them to a CSV called gguf_trending_models.csv with the header model_id,pipeline_tag.
• Read the data in gguf_trending_models.csv. Use that data to update the pipeline_tag for all models in gguf_trending_models.csv using the update_metadata tool.

Limitations

• Agents make mistakes, especially on complex operations. Break it down into simple tasks with manual verification.
• The Hugging Face Hub API has rate limits. Avoid massive-scale operations.