Skip to main content
fastcrw
All providersProvider terms
fastCRW Pergamon
Developer

Hugging Face Hub

huggingface.co

Metadata for every public model, dataset and Space on the Hugging Face Hub - search, repository records, file trees, commit history, dataset schemas and row previews, papers, collections and community profiles.

Showing the catalogue
Runs on your machine under the source's own terms

The crw MCP on your computer calls Hugging Face Hub directly. Calls cost no credits and the data does not pass through fastCRW. Install it once for Claude Code, Cursor, Codex, Windsurf, Gemini CLI or OpenCode, with your fastCRW API key in place of the placeholder:

Ask with this source

A prompt for your agent that names this source and how to call it through the crw MCP.

  • Collection details

    Use Hugging Face Hub (huggingface) from fastCRW Pergamon, capability "Collection details" (collection). You have a collection slug and want every repository or paper it contains. Example options: namespace=google, slug=gemma-4-69ce8ad93186d46744cb42f1. Use the crw MCP tools: crw_find_data to confirm the source and its options, then crw_get_data with provider "huggingface" and capability "collection". Tell me the credits used and cite the licence from the response. If the crw MCP tools are missing, install them with: npx -y crw-mcp install --api-key <your fastCRW API key>

  • Collection search

    Use Hugging Face Hub (huggingface) from fastCRW Pergamon, capability "Collection search" (collections). You want curated groupings of related repositories - a model family, a release, a reading list - rather than raw search results. Example options: owner=google, item=models/google/gemma-7b, q=gemma. Use the crw MCP tools: crw_find_data to confirm the source and its options, then crw_get_data with provider "huggingface" and capability "collections". Tell me the credits used and cite the licence from the response. If the crw MCP tools are missing, install them with: npx -y crw-mcp install --api-key <your fastCRW API key>

  • Daily papers

    Use Hugging Face Hub (huggingface) from fastCRW Pergamon, capability "Daily papers" (daily-papers). You want the ML papers the Hub community surfaced recently, with their Hub discussion and linked models. Example options: date=2026-10-15. Use the crw MCP tools: crw_find_data to confirm the source and its options, then crw_get_data with provider "huggingface" and capability "daily-papers". Tell me the credits used and cite the licence from the response. If the crw MCP tools are missing, install them with: npx -y crw-mcp install --api-key <your fastCRW API key>

  • Dataset details

    Use Hugging Face Hub (huggingface) from fastCRW Pergamon, capability "Dataset details" (dataset). You have a dataset's namespace and name and want everything the Hub knows about that repository. Example options: namespace=stanfordnlp, name=imdb, expand=cardData. Use the crw MCP tools: crw_find_data to confirm the source and its options, then crw_get_data with provider "huggingface" and capability "dataset". Tell me the credits used and cite the licence from the response. If the crw MCP tools are missing, install them with: npx -y crw-mcp install --api-key <your fastCRW API key>