--- title: modelread MCP emoji: 🧮 colorFrom: indigo colorTo: green sdk: gradio sdk_version: 6.32.0 app_file: app.py pinned: false license: other short_description: MCP server - will this HF model run on my hardware? tags: - mcp-server - mcp - local-llm - vram-calculator - quantization - gguf - llama-cpp - vllm - mlx - hardware --- # modelread MCP Will this Hugging Face model run on your GPU, Mac, phone, multi-GPU node or cloud instance? How fast, with which quant, and with what command? modelread reads the repository's real files and tensor headers, and returns a fit verdict, tokens/sec as a calibrated range, the longest context that fits, and a llama.cpp / vLLM / MLX launch line. **Tools:** `estimate_fit`, `compare_hardware`, `find_models`, `list_hardware`, `search_models` (read-only, deterministic). **Use it from an agent** - Hugging Face: add this Space at [huggingface.co/settings/mcp](https://huggingface.co/settings/mcp). - Any MCP client: `https://patrickbdevaney-modelread-mcp.hf.space/gradio_api/mcp/` - Or the main endpoint: `https://modelread.com/mcp` (Claude Code: `claude mcp add --transport http modelread https://modelread.com/mcp`) Website, model pages and methodology: https://modelread.com