Hugging Face Inc., the operator of a popular platform for hosting artificial intelligence models, today announced that it has closed a $100 million funding round led by Lux Capital.
Lux Capital was joined by Sequoia Capital, Coatue and more than a half dozen existing Hugging Face backers. The New York-based startup is now reportedly valued at $2 billion.
Hugging Face operates a platform called the Hugging Face Hub that hosts more than 100,000 open-source AI models. Developers can incorporate the AI models into software projects and submit their own neural networks to the platform. Besides neural networks, Hugging Face Hub also hosts related technical resources such as machine learning training datasets.
Hugging Face’s other open-source endeavors include managing ab AI research initiative known as BigScience. As part of the initiative, 900 researchers from around the world have developed what is being described as the world’s largest open-source natural language processing model. It’s currently being trained on a dataset that includes 300 billion words written in 46 languages.
Hugging Face says that more than 100,000 organizations rely on the Hugging Face Hub to support their AI projects. Among the platform’s users are tech giants such as Amazon.com Inc., Microsoft Corp. and Google LLC. The startup generates revenue by offering paid products to the organizations that use the platform.
Some of Hugging Face’s paid products focus on streamlining the AI development process. The startup offers an enterprise version of its AI hosting platform that companies can use to manage the code and datasets they create as part of their internal machine learning projects. Hugging Face also provides AutoTrain, which tasks the process of training AI models.
The other products offered by the startup focus on helping companies more easily manage their AI models. One such product, Inference API, enables companies to host neural networks in the cloud without having to manage infrastructure. Another Hugging Face offering, Infinity, can be used to increase the speed with which a neural network processes data in production.
Hugging Face reportedly generated less than $10 million from its portfolio of commercial offerings last year. The startup’s newly closed funding round could help it increase this number. Hugging Face plans to invest the new funding in product development, as well as research and open-source initiatives.
Hugging Face is the latest in a series of AI tooling startups to have raised funding at unicorn valuations over the past year. Dataiku Inc., which helps enterprise build and manage AI applications, last August raised $400 million from investors at a $4.6 billion valuation. AI development specialist DataRobot Inc. is worth $6.3 billion thanks to a $300 million funding round that it announced last June.
AI tooling providers’ growing valuations are partly the result of the increasing use of machine learning in enterprise software projects. This trend has also been reflected in the earnings reports of Nvidia Corp., the top maker of graphics processing units for training and running AI models. Nvidia’s revenues jumped 54% year-over-year, to $7.64 billion, last quarter.
The large number of startups competing in the AI tooling ecosystem could potentially lead to consolidation over the next few years. With $100 million in new funding now on its books, Hugging Face now has additional resources to make strategic acquisitions.