Switzerland is making a bold move onto the global artificial intelligence (AI) stage with the release of its first open-source, multilingual large language model (LLM). This initiative, spearheaded by leading Swiss institutions like ETH Zurich and EPFL, signifies a commitment to AI sovereignty, transparency, and linguistic inclusivity. Scheduled for release by the end of summer 2025, the model is poised to become a valuable resource for researchers, businesses, and public sector organizations worldwide.
The Swiss AI model distinguishes itself through its dedication to open-source principles. By providing unrestricted access to the model's code, weights, and training data, Switzerland is fostering a collaborative environment that encourages innovation and responsible AI development. This approach contrasts sharply with the proprietary models developed by tech giants, which often restrict access and limit the potential for customization and scrutiny. The open-source nature of the Swiss model promotes trust, regulatory compliance, and collaborative innovation.
Multilingualism is another key feature of this LLM. Trained on a vast dataset encompassing over 1,500 languages, the model is designed to be globally applicable. While English constitutes a significant portion of the training data, the project emphasizes linguistic diversity, making it particularly useful for translation, localization, and public sector communication across different languages. Approximately 40% of the training data covers non-English languages, reinforcing the model's usability across cultural and linguistic contexts. This focus addresses a significant gap in existing AI models, which often prioritize English and other major languages.
The development of this LLM is a flagship project of the Swiss AI Initiative, which was launched in December 2023. The initiative is supported by over ten academic institutions and boasts a team of more than 800 researchers. The Swiss National AI Institute (SNAI) was created by ETH Zurich and EPFL in October 2024 to provide a long-term and national perspective on AI-based research, education, and innovation and spearheads research on large-scale AI. The model is trained on the "Alps" supercomputer at the Swiss National Supercomputing Centre (CSCS) in Lugano. This high-performance infrastructure, powered by carbon-neutral electricity, ensures both technical excellence and environmental sustainability. The Alps supercomputer is equipped with over 10,000 NVIDIA Grace Hopper Superchips, making it one of the world's most advanced AI platforms. The Swiss AI Initiative is the largest open science/open source effort for AI foundation models worldwide. The initiative distributes 10-20 million GPU hours in 2025 for disruptive research projects through open calls. Researchers outside of Switzerland are encouraged to apply if they team up with at least one of the initiative's principal investigators and aim to create novel open science artifacts that benefit the Swiss, European, or global ecosystem and societal context.
Switzerland's commitment to AI extends beyond technological development. The initiative underscores the nation's vision for AI rooted in openness, accountability, and legal responsibility. By offering open access to the model's weights, code, and documentation, Switzerland asserts its vision for artificial intelligence. This stance aligns with the growing global emphasis on ethical AI development and the need for transparency and fairness in AI systems. Switzerland has always placed great importance on ethics, focusing on privacy, fairness, and accountability. The Swiss Federal Council has established an interdepartmental working group on AI, which adopted AI guidelines at the end of November 2020.
The release of Switzerland's open-source, multilingual AI model marks a significant step towards democratizing AI technology and promoting responsible innovation. By prioritizing transparency, accessibility, and ethical considerations, Switzerland is positioning itself as a key player in the global AI landscape. This initiative not only strengthens Switzerland's AI ecosystem but also contributes to the development of AI for the public good, fostering a future where AI benefits all of humanity.
