Overview: mcp-hfspace MCP Server 🤗

Title

mcp-hfspace MCP Server – A seamless integration tool for connecting to Hugging Face Spaces with minimal setup.

Installation:
- Install NodeJS and add the server configuration to claude_desktop_config.json.
- Use the NPM package @llmindset/mcp-hfspace.
- Ensure Claude Desktop version 0.78 or higher.
Basic Setup:
- Specify Hugging Face Spaces in the arguments.
- Set a working directory for file handling using --work-dir or MCP_HF_WORK_DIR.
- For private spaces, provide a Hugging Face token via --hf-token or HF_TOKEN.
Running the Server:
- Multiple instances can be run with different configurations.
- Supports both Claude Desktop Mode (default) and Base64-encoded resource mode.

Easy Integration: Connect to Hugging Face Spaces effortlessly.
File Handling: Supports upload/download of images, audio, and other files.
Multi-Space Support: Use multiple Spaces simultaneously (e.g., image generation, vision models, text-to-speech).
Private Space Access: Securely use private Hugging Face Spaces with a token.
Claude Desktop Mode: Optimized for Claude Desktop, with images embedded in responses and other files saved locally.
Prompt Automation: Auto-generates prompts for Spaces, improving usability.

Image Generation: Compare outputs from different models (e.g., shuttleai/shuttle-3.1-aesthetic).
Vision Models: Analyze images with models like merve/paligemma2-vqav2.
Text-to-Speech & Speech-to-Text: Convert text to audio or transcribe speech using spaces like parler-tts/parler_tts.
Chat & Reasoning: Engage with AI models like Qwen/Qwen2.5-72B-Instruct for complex queries.
Image-to-Image Processing: Annotate and analyze images with tools like microsoft/OmniParser.
Interactive Applications: Use models like DawnC/PawMatchAI for dynamic recommendations.

For more details, visit the introduction page.