MCP.so
Sign In

mcp-hfspace MCP Server 🤗

@evalstate

RepositoryHomepage

About mcp-hfspace MCP Server 🤗

MCP Server to Use HuggingFace spaces, easy configuration and Claude Desktop mode.

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "mcp-hfspace": {
      "command": "npx",
      "args": [
        "-y",
        "@llmindset/mcp-hfspace"
      ]
    }
  }
}

Tools

3

A list of available file and resources. If the User requests things like 'most recent image' or 'the audio' use this tool to identify the intended resource.This tool returns 'resource uri', 'name', 'size', 'last modified' and 'mime type' in a markdown table

Use semantic search to find an endpoint on the `Hugging Face Spaces` service. The search term will usually be 3-7 words describing a task or activity the Person is trying to accomplish. The results are returned in a markdown table. Present all results to the Person. Await specific guidance from the Person before making further Tool calls.

Call the FLUX.1-schnell endpoint /infer

Overview

What is mcp-hfspace?

mcp-hfspace is an MCP server that connects Claude Desktop and other MCP clients to Hugging Face Spaces, enabling access to thousands of AI models and applications. It is designed for developers and researchers who want to integrate Hugging Face’s AI capabilities into their workflows with minimal setup.

How to use mcp-hfspace?

Install NodeJS, then add the server to your claude_desktop_config.json with the command npx -y @llmindset/mcp-hfspace. Supply a list of Hugging Face Space names as arguments. Optionally set a working directory via --work-dir or MCP_HF_WORK_DIR, and supply a Hugging Face token for private spaces via --hf-token or HF_TOKEN.

Key features of mcp-hfspace

  • Connects to any Hugging Face Space automatically
  • Default space: black-forest-labs/FLUX.1-schnell for image generation
  • Supports file upload, download, and URL inputs
  • Private spaces supported with Hugging Face token
  • Multiple server instances for different tokens/working directories
  • Claude Desktop Mode with embedded image returns

Use cases of mcp-hfspace

  • Generate and compare images using different models (e.g., FLUX.1-schnell vs shuttle-3.1-aesthetic)
  • Use vision models to analyze images from files or URLs
  • Convert text to speech or transcribe audio with speech-to-text spaces
  • Perform image-to-image tasks like background removal and annotation
  • Run chat models like Qwen for reasoning tasks

FAQ from mcp-hfspace

What has replaced this project?

The project has been superceded by the official Hugging Face MCP Server and Gradio MCP Endpoints. Alternatively, you can run hf-mcp-server locally as a STDIO server with SSE support.

What are the runtime requirements?

Requires NodeJS and Claude Desktop 0.78 or greater. The server is installed via npm as @llmindset/mcp-hfspace.

How does file handling work?

In Claude Desktop Mode, images are returned in tool responses; other files are saved in the working directory with their path returned. URLs can also be supplied as inputs.

What limitations exist?

Endpoints with unnamed parameters are unsupported. Some complex Python types may not fully translate to MCP formats. Claude Desktop may timeout on long jobs (hard 60s limit). Check the working directory for results produced after timeout.

How do I configure private spaces and tokens?

Supply a Hugging Face token via --hf-token=hf_... or the HF_TOKEN environment variable. Private spaces are supported, and the token is used to download generated content.

Frequently asked questions

What has replaced this project?

The project has been superceded by the official Hugging Face MCP Server and Gradio MCP Endpoints. Alternatively, you can run `hf-mcp-server` locally as a STDIO server with SSE support.

What are the runtime requirements?

Requires NodeJS and Claude Desktop 0.78 or greater. The server is installed via npm as `@llmindset/mcp-hfspace`.

How does file handling work?

In Claude Desktop Mode, images are returned in tool responses; other files are saved in the working directory with their path returned. URLs can also be supplied as inputs.

What limitations exist?

Endpoints with unnamed parameters are unsupported. Some complex Python types may not fully translate to MCP formats. Claude Desktop may timeout on long jobs (hard 60s limit). Check the working directory for results produced after timeout.

How do I configure private spaces and tokens?

Supply a Hugging Face token via `--hf-token=hf_...` or the `HF_TOKEN` environment variable. Private spaces are supported, and the token is used to download generated content.

Comments

Basic information

Category

AI & Agents

License

MIT

Transports

stdio

Author

evalstate

More AI & Agents MCP servers

21st.dev Magic AI Agent

21st-dev

It's like v0 but in your Cursor/WindSurf/Cline. 21st dev Magic MCP server for working with your frontend like Magic

5.3K

Orders of Magnitude — x402 API Catalog

OrdersOfMagnitudeLLC

1000+ pay-per-call API endpoints across finance, weather, geography, economic data, agent infrastructure (memory, coordination, secrets, simulation, verification), and NS compute. Payments via x402 protocol on Base/USDC.

q-ring

I4cTime

Quantum-inspired keyring for AI coding agents. Secure secrets with superposition, entanglement, tunneling, and teleportation — CLI + 44 MCP tools.

3

Nero AI Image Processing

nero-com

We provide AI-powered tools for image enhancement, upscaling, background removal, photo restoration, colorization, denoising, and compression. All features are available through our API documentation.

🛡️ A.I.G（AI-Infra-Guard）

Tencent

A full-stack AI Red Teaming platform securing AI ecosystems via OpenClaw Security Scan, Agent Scan, Skills Scan, MCP scan, AI Infra scan and LLM jailbreak evaluation.

4.1K