🤗📡🤝
The Public AI Inference Utility is now an official inference provider on Hugging Face, bringing public AI to developers worldwide.
September 16, 2025
Through this collaboration, users can now access and test public and sovereign AI models like Switzerland's Apertus and Singapore's SEA-LION directly through the developer-friendly Hugging Face ecosystem. This is an important step for making public AI more accessible to developers, researchers, and organizations worldwide.
- Try our models on HuggingFace (more to come soon!)
- See our organization card on HuggingFace
- Learn more about our infrastructure
🙏 to Julien (pictured above) and Simon (pictured below) from HF, who were super helpful & responsive in setting this up! The entire process was surprisingly easy; we just made a few PRs directly to the HF site. Also hat tip to Poli (not pictured) who made the connection.
What this means for developers
As an official Hugging Face inference provider, Public AI now offers:
- Direct API access to public andsovereign models through HuggingFace's infrastructure
- Seamless integration with existing Hugging Face workflows and tools
- Global availability with our distributed inference network
- Transparent pricing (currently free during Swiss AI Weeks!) and usage tracking through Hugging Face's platform
This partnership makes it easier than ever for developers to build applications using public AI models, whether you're prototyping a new idea or deploying at scale.
Getting started
Ready to try public AI on HuggingFace? Here's how to get started:
- Browse our models: Visit our HuggingFace organization to see available models
- API access: Use HuggingFace's inference API with our models
- Documentation: Check out our platform docs for detailed integration guides
The future of public AI
This partnership is just the beginning. We're committed to expanding our model offerings on Hugging Face, including future releases from our partner institutions at national labs and national champions.