A New Kind Of AI Built For The Public Good.

In a world increasingly shaped by artificial intelligence, who controls the tools, and whose values they reflect, matters more than ever. That’s why ETH Zurich, in collaboration with EPFL and the Swiss National Supercomputing Centre, is making waves with a bold new initiative: the launch of a large language model (LLM) built entirely on public infrastructure, trained with transparency, and released for the benefit of all.

Instead of keeping the inner workings locked behind corporate walls, this new model will be fully open-source. That means researchers, educators, governments, and even startups will have access to everything—from the code and training data to the model itself. It’s designed to be not just powerful, but accountable, offering an alternative to today’s closed, proprietary AI systems.

But it’s not just about openness, it’s about inclusivity. The model is being trained on content in over a thousand languages, making it one of the most globally accessible AI systems ever built. While many tools today are heavily English-dominant, this one embraces the full spectrum of human expression, amplifying underrepresented voices in the digital world.

And there’s more: the entire project runs on a supercomputer powered by carbon-neutral energy, proving that technological innovation doesn’t have to come at the planet’s expense.

At a time when trust in tech is wavering, this effort signals a new kind of leadership - one that puts public benefit, transparency, and global equity at the center of AI’s evolution.

This isn’t just a language model. It’s a powerful step toward AI that truly serves the people.