HELSINKI — A new, transformative collaboration is underway in Europe as some of the continent’s leading AI companies and research institutions unite for the OpenEuroLLM project. This unprecedented initiative brings together 20 prominent European entities to develop open-source, multilingual language models that will be crucial for commercial, industrial, and public service applications. The project is co-led by Peter Sarlin from AMD Silo AI in Finland and Jan Hajič from Charles University in Czechia, with the aim of creating transparent, high-performance AI models that will help European companies and public organizations remain competitive globally.

The OpenEuroLLM project is a direct response to the need for greater digital sovereignty and competitiveness within Europe. By developing these open-source models, the consortium hopes to democratize access to cutting-edge AI technologies, ensuring that European companies are equipped to thrive in the global market and that public organizations can deliver impactful services to citizens. The initiative emphasizes the importance of openness, transparency, and community involvement, reflecting the core values of the European tech ecosystem.

Working closely with open-source and open science communities such as LAION, open-sci, and OpenML, the project ensures that the models, data, and software will remain fully open. This approach allows the models to be fine-tuned and adapted for specific industrial and public sector needs, fostering linguistic and cultural diversity in AI development. This will enable European businesses to create high-quality AI-driven products and services, with special attention to preserving Europe’s rich cultural heritage.

Funded by the European Commission under the Digital Europe Programme, the project has earned the STEP (Strategic Technologies for Europe Platform) seal and builds on the success of prior European AI projects. The consortium’s collective experience, including access to extensive high-quality data repositories and pilot language models, provides a strong foundation for the ambitious work ahead. The OpenEuroLLM project officially kicks off on February 1st, 2025, with the collaborative efforts of academic, industrial, and supercomputing centers in Europe.

Full List of Partners:
Universities, Research, and Public Organizations:
Charles University (Institute of Formal and Applied Linguistics), Czechia (Coordinator)
Alliance for Language Technologies EDIC (ALT-EDIC), France
Eindhoven University of Technology, the Netherlands
ELLIS Institute Tübingen, Germany
Fraunhofer IAIS, Germany
Lindholmen Science Park (AI Sweden), Sweden
Research Center Juelich, Germany
University of Helsinki, Finland
University of Oslo, Norway
University of Turku, Finland
University of Tübingen (Tübingen AI Center), Germany

Companies:
Silo GenAI (AMD Silo AI), Finland (Co-lead)
Aleph Alpha Research, Germany
ellamind, Germany
LightOn, France
Prompsit Language Engineering, Spain

EuroHPC Centres:
Barcelona Supercomputing Center, Spain
Cineca Interuniversity Consortium, Italy
CSC – IT Center for Science, Finland
SURF, the Netherlands