Hugging Face

Hugging Face is a platform where the machine learning community creates, discovers, hosts, and collaborates on models, datasets, and applications. It supports text, image, video, audio, and 3D modalities.

Users can browse more than 2 million models, 1 million applications in Spaces, and 500,000 datasets. The platform also includes open-source tools such as Transformers, Diffusers, and Safetensors.

Hugging Face offers paid compute and enterprise features, including Inference Endpoints, GPU access for Spaces, team security controls, and dedicated support. Enterprise plans start at $20 per user per month, while GPU compute starts at $0.60 per hour.

Highlights

• Host and collaborate on public models, datasets, and applications
• Browse models, datasets, and interactive Spaces
• Access 45,000+ models through a unified Inference Providers API
• Deploy models with Inference Endpoints or GPU-enabled Spaces
• Open-source libraries for transformers, diffusion models, and model storage