The National Science Foundation and NVIDIA will invest $152 million in the development of advanced, open-source artificial intelligence models aimed at accelerating American scientific discovery. The public-private investment will support the Open Multimodal AI Infrastructure to Accelerate Science project led by the Allen Institute for AI, or Ai2, NSF said Thursday.
Table of Contents
Public-Private Investment for Open Source AI Models
NSF will provide $75 million, with NVIDIA contributing $77 million. The initiative supports the White House AI Action Plan and aims to ensure the United States remains a leader in AI-powered research and innovation.
“As called for in the AI Action Plan, the Trump administration is committed to ensuring America fosters the leading open-source and open-weight models that set the global standard for business and academic research worldwide,” said Michael Kratsios, director of the White House Office of Science and Technology Policy and a two-time Wash100 Award recipient. “This partnership between NSF and NVIDIA puts the Action Plan to work and demonstrates the power of the American innovation ecosystem. We look forward to the novel research and scientific breakthroughs to come through open-source model development and AI-enabled science.”
The investment is expected to address the budget constraints surrounding the creation and research of powerful AI models, which academic researchers use to explore various topics.
Boosting Research Across Disciplines
With the funding, Ai2 will create open-source, multimodal large language models trained on scientific data to help researchers analyze literature, generate code and visualization, and connect new insights to past discoveries to accelerate breakthroughs in fields such as materials science, biology and energy. The project also includes workforce development to expand AI expertise beyond traditional tech hubs.
In addition to Ai2, teams from the University of Washington, the University of Hawaii at Hilo, the University of New Hampshire and the University of New Mexico will also receive support from NSF and NVIDIA.
NVIDIA CEO Jensen Huang called AI “the engine of modern science,” while Ai2 CEO Ali Farhadi stressed the need for open collaboration to maintain U.S. scientific leadership.
Initial applications of the open-source AI models are expected to advance material discovery, improve protein function prediction for biomedical advancements and address weaknesses in current LLMs.
Related Articles
The General Services Administration has launched USAi, a secure generative artificial intelligence suite designed to help federal agencies experiment with AI tools and accelerate AI adoption. GSA said Thursday the launch of USAi advances the priorities in the White House’s America’s AI Action Plan, which seeks to strengthen U.S. leadership in AI through coordinated federal action, streamlined adoption and smarter infrastructure. A panel discussion at the Potomac Officers Club’s 2025 Navy Summit on Aug. 26 will explore generative AI and how it optimizes decision-making within the service branch. Book your spot now at this GovCon networking event! Expanding Federal Government’s
Nextgov/FCW reported that the Federal Acquisition Regulatory Council on Thursday released new model deviation text for six parts of the FAR as part of the Revolutionary FAR Overhaul, or RFO, initiative. In April, President Donald Trump signed an executive order directing his administration to amend FAR to streamline the federal procurement process and eliminate barriers to doing business with the government. In June, the FAR Council released model deviation text for sections related to emergency acquisitions, contract modifications and acquisition of information and communication technology, or ICT. The overhauled FAR parts include sections related to administrative and information matters; required sources of
An upcoming strategy will accelerate the process by which the Department of the Navy can extract insights from data and transform them into new tactics and technologies, according to Stuart Wagner, chief data and artificial intelligence officer for the DON. At a recent event, the official shared that the final draft of the department’s data and AI weaponization strategy will be signed “in the next month or so.” He said that the document will identify six broad goals, which will include rapidly transitioning AI systems from pilot to operational, equipping personnel with data and AI skills, and facilitating collaboration with