Arcee AI Partners with US Department of Energy to Build Open-Weight Scientific AI Model

Arcee AI, a US based developer of large language models, announced a partnership with the US Department of Energy (DOE) to build Genesis Science 1 (GS1). The announcement was made via Arcee AI's official account on X on Friday, July 24, 2026. GS1 is described as an American open weight AI model and a governed research harness designed to complete scientific computing workflows while maintaining a reproducible record of its work. The model will be shaped by scientists with deep understanding of scientific work. The DOE has opened a contributor program for researchers, laboratories, universities, companies, and nonprofit organizations to participate in GS1's development. Arcee AI is also in discussions with infrastructure partners that could add training or evaluation capacity.

The partnership is part of the Genesis Mission, a national DOE initiative to build the world's most powerful scientific platform to accelerate discovery science, strengthen national security, and drive energy innovation. The Genesis Mission was first announced by the DOE in October 2025. In December 2025, the DOE announced collaboration agreements with 24 organizations interested in advancing the Genesis Mission. This step marked an early commitment by the US government to engage the private sector and academia in a large scale AI project for science. Mark McQuade, founder and CEO of Arcee AI, called the partnership a big day for Arcee and for American open weight AI.

In a LinkedIn post, he expressed pride in partnering with the DOE to build the next generation of open models for scientific research. GS1 differs from closed commercial AI models because its model weights will be open, allowing researchers to inspect, modify, and reproduce the model's work. This approach aligns with the principles of open science, which emphasize transparency and reproducibility. The model is designed to handle complex scientific computing workflows, such as physics simulations, genomic data analysis, or climate modeling. By maintaining a reproducible record, GS1 enables other researchers to verify and build upon its results. The DOE's contributor program aims to gather input from various stakeholders in the US research ecosystem.

DOE national laboratories, research universities, technology companies, and nonprofit organizations are invited to contribute to GS1's development. Arcee AI stated that much work remains and invited public participation to build openly, starting with GS1. The company has not disclosed a release timeline or detailed technical specifications for the model. The partnership marks a significant step in the US government's efforts to develop domestic AI capacity for scientific research. By involving the private sector and research community, the DOE hopes to accelerate scientific discovery across multiple fields. The Genesis Mission itself is an ambitious initiative aimed at building the world's most powerful AI platform for science. The initiative includes the development of AI models, computing infrastructure, and cross sector collaboration.

The DOE has long been a leader in high performance computing through its national laboratories. With the Genesis Mission, the DOE seeks to integrate AI into scientific workflows to accelerate discoveries that previously took years. Arcee AI, founded in 2023, focuses on developing small, efficient, and customizable language models for specific applications. This partnership places Arcee AI at the forefront of AI development for science. No information has been released regarding the budget or contract value of this partnership. Arcee AI and the DOE have not issued further official statements beyond the social media announcement. Industry observers note that the openness of the GS1 model could set a precedent for future government AI projects.

If successful, GS1 could become a model for open AI collaboration between government, industry, and academia. The development of GS1 also has the potential to influence AI policy direction in the US, particularly regarding standards for transparency and reproducibility in government funded research. The DOE appears to be choosing an open approach to maximize scientific and national security impact. This move also reflects a global trend toward more open and auditable AI, especially for applications in science and national security. Other countries, such as the UK and Japan, have launched similar initiatives to develop AI for science. Arcee AI and the DOE plan to continue updating the public on GS1's progress through the contributor program and other communication channels.

The research and industry communities are awaiting further details on technical specifications and the model's development timeline.