by Hugging Face
Hugging Face is a platform for hosting and sharing machine learning models, datasets and demo apps. It also offers inference services and open-source libraries.
Web appMCP serverFreemium
Add this server URL to an MCP-compatible client.
Step 1: Create an account and a model, dataset or Space repository on the Hub.
Step 2: Push files with Git or the Hub Python library or CLI.
Step 3: Create a Space to share a demo built with Gradio or Docker.
Step 4: Call hosted models through Inference Providers, or deploy your own on Inference Endpoints.
Step 5: Upgrade to PRO, Team or Enterprise for more storage, quota and organization features.
Models, datasets, Spaces and Buckets can be stored on the Hub. Public and private repositories are both supported.
Inference Providers let you call models hosted by partner providers. Inference Endpoints deploy models on dedicated, fully managed infrastructure.
PRO adds ZeroGPU quota, highest queue priority and Spaces Dev Mode, which allows fast iteration through SSH or VS Code.
The docs include guides for deploying Hugging Face models on AWS, Microsoft Azure and Google Cloud.
Environment variables
Hugging Face DatasetsHugging Face Datasets is the Hub's browsable index of machine-learning datasets, with full-text search and filters. It is for practitioners looking for datasets by task, modality, format or license.
MasaMasa is a real-time data network for AI agents and applications, with data from sources such as X, Discord, Telegram, podcasts and YouTube. People can also run miners to contribute data and earn token rewards.
Sahara AISahara AI builds AI agents, data services and a blockchain for AI assets. Its products include the Sorin trading agent, a data services platform and enterprise agent solutions.