Hugging Face is an AI infrastructure platform and community hub for open-source machine learning, hosting models, datasets, and interactive demo Spaces. Clément Delangue, Julien Chaumond, and Thomas Wolf founded the company in New York City in 2016, originally as a teen-focused chatbot app, before pivoting toward the open-source ML tooling that now defines it.
Its Transformers library became close to a default choice for working with NLP and multimodal models in Python, and the public Hub, which stays free, functions as the open corpus and community layer that anchors the business; monetization happens around the edges through Inference Endpoints, PRO subscriptions, and Enterprise features. Dedicated GPU endpoints bill by the hour, from CPU instances at $0.033/hour up to H100 GPUs at $10/hour.
The platform serves ML researchers, individual developers, and enterprise AI teams that need model hosting and deployment without building infrastructure from scratch. The Hub itself remains free, PRO access for individuals runs $9/month, Team access runs $20/user/month, and Enterprise runs $50/user/month, each bundling Inference Providers credit.
