Switzerland unveils Apertus, an open source national LLM developed by ETH Zurich, EPFL, and CSCS to strengthen sovereign AI infrastructure and reduce reliance on US and Chinese models.
Switzerland has entered the global AI race with the launch of Apertus, a national Large Language Model (LLM) designed to establish sovereign AI infrastructure as an alternative to American and Chinese dominance.
Developed by researchers and engineers from ETH Zurich, EPFL, and the Swiss National Supercomputing Centre (CSCS), Apertus is built as a fully open source system. Its name, derived from the Latin word for ‘open’, reflects its transparent architecture. The release includes training data, model weights, and intermediate checkpoints, making it one of the few LLMs at this scale to embrace open principles.
Apertus is available in two configurations: an 8-billion-parameter version for individual use and a 70-billion-parameter version for large-scale applications. The model is documented and accessible on Hugging Face, where it can be used as a foundation for developers to build applications such as chatbots, translation systems, and educational tools. At present, Swisscom business customers are the only users able to access the model via its AI platform.
“We see it as a driver of innovation and a means of strengthening AI expertise across research, society, and industry,” said Thomas Schulthess, Director of CSCS and Professor at ETH Zurich.
“Apertus is built for the public good. It stands among the few fully open LLMs at this scale and is the first of its kind to embody multilingualism, transparency, and compliance as foundational design principles,” added Imanol Schlag, Technical Lead of the LLM Project and Research Scientist at ETH Zurich.
Looking ahead, the developers announced that Swiss AI Weeks hackathons will provide the first opportunity for developers to experiment with Apertus, test its capabilities, and provide feedback for future improvements.
