Hugging Face Launches Project ProtoStar to Make Publicly Funded Research AI-Ready, First Climate Challenge Opens November 2

On October 29, 2026, Hugging Face announced Project ProtoStar, an initiative to make datasets, models and code from publicly funded research easier to discover and reuse, especially for AI-driven research. Participants commit to releasing suitable scientific artifacts in AI-ready, machine-actionable formats and to show the value of reuse through a series of open challenges. The first, hosted by Carnegie Mellon University, opens on November 2.

The founding partners are Carnegie Mellon University, Purdue University, the Georgia Institute of Technology, Arizona State University and The George Washington University, working through the Open Forum for AI (OFAI) as university coalition partner, together with the Astera Institute. Hugging Face is the infrastructure partner, and the Open Research Community Accelerator (ORCA), through its Fostering Innovation Network, is the community partner. Participating organizations formalize their involvement in a memorandum of understanding that supports releasing suitable scientific datasets, models and code in AI-ready formats on open infrastructure. Organizations that want to join can write to [email protected].

The post's starting point is that making research public does not make it usable. The U.S. federal government invests about $190 billion a year in research and development, agencies such as the National Institutes of Health (NIH) and the National Science Foundation (NSF) have strengthened public-access requirements and data-sharing plans, and Hugging Face is among ten generalist repositories NIH lists. Yet scientific datasets, models and code are often scattered across incompatible systems, kept in static or legacy formats, or released without the documentation, metadata and provenance needed to reuse them. ProtoStar aims to close that gap by encouraging systematic sharing in AI-ready formats, backed by open infrastructure and practical publishing guidance.

The post cites Hugging Face's own statistics as evidence of demand. Hugging Face hosts more than 17,000 public datasets across core scientific domains such as genomics, protein design, chemistry, weather and climate, Earth observation and astronomy, close to 29,000 when medicine and mathematics are included, with a comparable number of scientific models. Researchers shared more than three times as many new science datasets in the first nine months of 2026 as in the same period of 2025, and growth stayed substantial after excluding the largest single uploader. The 18 public repositories that NIH's National Center for Biotechnology Information (NCBI) maintains on the platform have received more than 11 million downloads, NASA IMPACT's 27 repositories about 770,000 as of October 26, 2026, and 90% of dataset downloads on Hugging Face are of fully documented datasets. These are the platform's own figures, not an independent assessment.

The first challenge, Climate Risk Intelligence, is hosted by Carnegie Mellon University from November 2 to 13, 2026. Participants use climate records through 2018 for 210 U.S. cities to predict actual conditions from 2019 to 2026, so models are judged against observed weather rather than hypothetical scenarios. They forecast four indicators: average annual temperature, total annual precipitation, the hottest day of the year, and cooling degree days, a measure of air-conditioning demand. Predictions are compared with simple statistical baselines such as historical averages and linear trends, to see whether AI approaches add meaningful accuracy. Participants also propose a short climate adaptation policy idea, which is judged separately. Teams submit a Hugging Face dataset, forecasting is scored on a public leaderboard by prediction error against the baselines, and final results come in early 2027 once the complete 2026 weather records exist. One inconsistency: the announcement says the forecast period is 2019 to 2026, while the challenge page itself says 2019 to 2025. This article does not pick between them.

Keep in mind that this is a newly announced initiative, not research with results. The post sets no quantitative standard for "AI-ready" and does not say how many institutions will actually release data. Its test will be whether the challenges produce concrete examples of reuse and expose the barriers in documentation and interoperability. For research organizations, the practical steps today are to contact the organizers or to check whether the data they plan to release has full documentation and metadata.