Apertus is an open source Large Language Model (LLM) developed in Switzerland. This documentation shows you how to get started with the LLM, whether as user, researcher, or advanced contributor: we are maintaining this knowledge base for you, and could ✉️ use your feedback.

  • For casual users: There are a number of apps in our Community Showcase, as well as a range of providers providing high quality inference services for Apertus listed on our Getting Started page.
  • For developers: The providers listed on the Getting Started page offer services in Apertus LLM hosting online for you to use via API. We test and verify each listing, but make no guarantees about their uptime or performance.
  • For self-hosting: You can download one of the Apertus model distributions for free from Hugging Face and deploy it locally into a desktop tool, such as the ones listed in our Deployment Guides. Check if you have enough system resources (e.g. 16 GB for BF16 versions, ~6 GB for quantized builds), or look for community-supported quantized releases.
  • For Apertus contributors: If you have questions or constructive feedback about your experience with Apertus, it is easy to engage with the developer and user community - just look for the official Swiss AI projects on Hugging Face, GitHub, and social media - see our Contact page for links on how to get involved.

This website is running on the Hugo static site engine and Lotus docs theme.

⬅️ Back to the home page

About the project

The model development team is part of the Swiss AI Initiative, which started in late 2023. This is a platform for over 80 data science projects including the LLM development. Key highlights of the LLM project, as announced in July, include:

  • Multilingualism: Trained on more than 15 trillion tokens across 1,500+ languages, 40% non-English - equal usage cost across languages - see @epfml
  • Performance: This is a large model (8 billion and 70 billion parameters), trained on a lot of tokens, and it will be continue to be actively optimized.
  • Open & Transparent: Published under Apache-2.0 license - including source code, weights, and open training data.
  • Data Privacy: Developed with attention to the GDPR, the EU AI Act, and Swiss data protection law - see Fan et al 2025
  • Infrastructure: Developed on the new Alps supercomputer at CSCS with over 10,000 NVIDIA GH200 Grace-Hopper chips
  • Global Reach: Research and borderless applications in mind, for sovereign and international public-interest AI.

Sources

For further information:

Edit this page