OpenEuroLLM Project Announces Development of Open-Source Multilingual AI Models

OpenEuroLLM Project, a European alliance tasked with developing open-source artificial intelligence (AI) models, was announced on Monday. The project is backed by the European Commission, which has also awarded it the Strategic Technologies for Europe Platform (STEP) Seal, recognising it as a critical technology project. The group aims to develop a family of multilingual large language models (LLMs) that can be proficient in all languages under the European Union (EU). These models will also be developed transparently and adhere to the regulatory framework of the bloc.

European Commission Backs the OpenEuroLLM Project

In a post on X (formerly known as Twitter), the official handle of the European Commission announced that the OpenEuroLLM Project has been awarded the first STEP Seal of the year. Notably, STEP Seal is a quality label awarded to high-quality projects meeting the minimum quality requirements under the Digital Europe Programme. The selected projects are provided visibility on the STEP platform and are promoted by the EU to attract investors easily.

In a press release, the OpenEuroLLM Project stated that it started the work to develop AI models on February 1. The project comprises a consortium of 20 European research institutions, companies, and EuroHPC centres coordinated by Jan Hajič from Charles University in Czechia and is co-led by Peter Sarlin, the Co-Founder and CVP at AMD Silo AI.

The OpenEuroLLM Project plans to build a family of high-performance multilingual LLMs that will be released to the open-source community for commercial, industrial, and public services. The project confirmed that it will adhere to the strict regulatory policies of the EU and will be transparent about data procurement.

Once the AI models have been made available, the project will also publicly release the documentation, training and testing code, and evaluation metrics of the AI models. This will be done to ensure that the LLMs can be fine-tuned and instruction-tuned for specific industry and public sector needs.

“The transparent and compliant open-source models will democratise access to high-quality AI technologies and strengthen the ability of European companies to compete on a global market and public organizations to produce impactful public services,” added the press release.

Notably, the European Commission has already funded the OpenEuroLLM project under the Digital Europe Programme, and it is expected to gain more investors in the coming weeks. So far, there is no roadmap on when these models will be released, or what would be the focus area of these models.