We are committed to developing and sharing cutting edge technology that focuses on improving AI in low-resource languages.
The members of our team have previously developed multilingual datasets and models such as Suzume, LB-Reranker, and Mitsu.
Now we are focusing on supporting even more languages through cutting edge research.
Our mission is to create and openly share the foundational tools—from robust datasets to state-of-the-art models—that enable technological progress for these communities.
We believe in empowering local developers and researchers by providing the open-source building blocks necessary for innovation in machine translation, content creation, and information access.
Follow us on Huggingface for dataset and model releases!