Closing the gap between academia and the AI frontier
Open Athena is a nonprofit that empowers academic labs to build scientific foundation models. We provide world-class engineering talent, compute, and coordination so that universities can remain central to discovery in the age of AI.
Open Athena collaborates with leading research institutions and their top scientists
.avif)
.avif)
Open Athena is generously supported by partners in compute and infrastructure
Distributed globally at hubs of AI and research
Open Athena maintains offices in New York, San Francisco, London, and Boston, hubs where frontier AI and frontier science intersect.
Questions we’re often asked
Why is Open Athena needed?
We believe in academia's role in the pursuit and open sharing of knowledge and understanding of the world. In the current era of artificial intelligence, academia does not have access to the resources and technical expertise to take promising ideas and proofs of concept to the frontier of their respective fields. Open Athena exists to help academic labs and open science take these nascent models and bring them to the next level.
How does Open Athena support projects?
We bring 4 things to bear:
- Top tier research scientists and infrastructure engineers with experience in building cutting-edge AI across many commercial and scientific domains.
- Compute resources and relationships with neoclouds and hyperscalers interested in supporting meaningful scientific research.
- Project coordination to help ensure we're working together as efficiently as possible.
- Curiosity and excitement about the opportunity to work with you to answer challenging and meaningful scientific questions.
What types of projects does Open Athena support?
Open Athena partners with academic labs to co-develop scientific foundation models that advance entire fields. Our process is simple, structured, and designed to amplify what you do best: science.
We’re looking to build highly-effective foundation models for advancing science. We are looking for collaborations where we have an opportunity to build state-of-the-art models that are effective across a wide range of tasks, and potentially have downstream users who may be interested in leveraging such a model were it to exist.
Labs should be collaborative, open to multi-university work, open to working in the open where possible, and committed to publishing and sharing results.
