Open ProtoStar challenges that put publicly funded research to work.
Project Protostar
community
AI & ML interests
None defined yet.
Recent Activity
Organization Card
Hugging Face and partners across science and research are working to make publicly funded datasets, models and code easier for researchers and AI systems to discover, use and build on.
ProtoStar is an initiative to turn publicly funded scientific research into resources anyone can put to work. It brings together open technical infrastructure, practical publishing support and community challenges to encourage the release and reuse of scientific datasets, models, code and documentation.
Access alone does not guarantee reuse.
Public access to scientific results is expanding, but data is still scattered across institutional portals, buried in PDFs, or published without the documentation, licensing and provenance needed to combine it with other work. Research on the barriers to AI-ready science shows they are institutional as well as technical: data curation needs better tools, support and incentives.
The demand is already there. The Hugging Face Hub hosts more than 17,000 public scientific datasets and a similar number of scientific models, and in the first nine months of 2026 researchers shared over three times as many new science datasets as in the same period of 2025. Documentation is what makes them count: 90% of dataset downloads on the Hub come from fully documented datasets.
How ProtoStar works
01
Share research in usable forms
Support researchers who want to release datasets, model weights, code and documentation in open, machine-readable formats, with clear metadata and instructions for reuse.
02
Build lasting homes for it
Help institutions maintain discoverable repositories on open infrastructure, including the Hugging Face Academic Hub, with licensing, versioning and preservation practices that fit their needs.
03
Put it to the test in public
Run challenges, hackathons and competitions with the open-science community that reuse those artifacts, surface barriers and show what better access can enable.
Participating organizations keep control over which materials are appropriate to release and how they are shared. ProtoStar builds on Hugging Face’s guidance for public organizations preparing data for AI, and complements national efforts to expand AI for science, including the Genesis Mission.
Founding partners
Working in the open
We are running this collaboration openly. Participating organizations have signed Memoranda of Understanding with Hugging Face: non-binding, non-exclusive commitments to release publicly funded research in AI-ready formats on open infrastructure.
There is a lot of work remaining to clear the hurdles in the system, from formats and documentation to incentives and infrastructure. This is the first step.
Read the signed MOUs ↗The universities take part through the Open Forum for AI, and ORCA through its Fostering Innovation Network. Hugging Face provides the open infrastructure that helps research travel beyond the institutions that produced it.
Carnegie Mellon is hosting the inaugural ProtoStar challenge: AI systems that analyze extreme weather trends and recommend climate adaptation strategies, built on CMU's historical weather records for 210 U.S. cities.
Using only data up to 2018, forecast each city's weather for 2019–2025, then propose one policy idea that helps cities cope with a climate risk. The challenge runs November 2–13, 2026 and is open to everyone. Prizes: $1,000, $500 and $250 in Hugging Face credits for the top three teams, plus $2,500 for the top CMU team.
Rules, data & registration ↗
ProtoStar welcomes universities, federal science agencies, national laboratories, nonprofits and other institutions interested in making publicly funded research more reusable.
Scientific research institutions
For federal agencies, national labs and nonprofit research institutes: a turnkey path to open, AI-ready release and public access compliance on modern, interoperable infrastructure, with an immediate line to the open research community.
Universities
Meet sponsor-mandated research release requirements by publishing on infrastructure for open science, and connect your output directly to the community on Hugging Science.
Researchers and developers
ORCA and OFAI run the open challenges and hackathons with Hugging Face and other open science organizations. Everyone is welcome to compete.
How institutions join
01
Get in touch
Email protostar@orcaopen.org with the research you would like to release and who will lead it.
02
Sign the MOU
A short, non-binding, non-exclusive Public Access MOU with Hugging Face. Read the ones partners have signed.
03
Open your home on the Hub
Set up your organization on Hugging Face and the Academic Hub, with a point of contact, clear licensing and metadata.
04
Publish and put it to work
Release datasets, models and code in AI-ready formats, then run a challenge with ORCA, OFAI and the open-science community.
Sources
Channing, G. & Ghosh, A. (2025). AI for Scientific Discovery is a Social Problem.arXiv:2509.06580 ↗
Ghosh, A. & Jernite, Y. (2025). Empowering Public Organizations: Preparing Your Data for the AI Era.Hugging Face Blog ↗
Hugging Face Hub data on public science datasets and models, as of October 19, 2026.hf-science-growth ↗
Science officer’s log · why “ProtoStar”?
A protostar is a star still gathering mass before it ignites — a fair description of a public science ecosystem that has the material but not yet the fusion. It is also the USS Protostar (NCC-76884), a small experimental ship crewed by people who were not supposed to be running a starship. Both felt about right.
Live long and publish openly. đź––
models 0
None public yet