Researching the next generation of multilingual AI
Okalai AI works at the frontier of Large Language Models (LLMs), pushing the boundaries where today’s LLMs fall short, especially in multilingual contexts. Originally launched in 2021 through a small series of AI events, Okalai AI became an independent research institute in 2025. We develop foundational methods that ensure language tech goes beyond high-resource settings.
Research
We bring underserved languages into multilingual NLP. Through original datasets, tools, and models, we expand the reach of language tech while contributing to the broader development of multilingual systems.
Education
We create educational materials for both people and machines. This includes resources that support both language learning, and the training of language models.
Impact
By working with real-world data, our projects can support practical applications in places currently not covered by mainstream language tech.