🧡Open Source Dutch AI

Dutch LLMs for Edge & Privacy

We post-train small language models in Dutch for edge inference and privacy-first applications. All our models are open source and optimized for local deployment.

Why openoranje?

We believe in accessible, privacy-respecting AI that speaks Dutch natively.

Privacy First
Run models locally without sending data to the cloud. Your data stays on your device.
Edge Optimized
Small, efficient models designed to run on consumer hardware and edge devices.
Native Dutch
Post-trained specifically for Dutch language understanding and generation.

Latest Releases

Our most recent models and research publications.

Modelv1.0
OpenOranje/TweeTaal-nl-en-0.6B
Our flagship 0.6B parameter model, fine-tuned for Dutch to English and English to Dutch translation.
0.6B params•November 2024

Join the Dutch AI Community

Contribute to our open source models, report issues, or just say hello. Together we're building the future of Dutch language AI.