Other languages
Models for African, Southeast Asian and other languages. 14 models on EveryLLM so far.
This group covers models for the many languages outside the big families: the languages of Southeast Asia, Africa, Latin America, Central Asia and the Pacific, as well as minority and indigenous languages. Most have little text online, so builders gather data from radio, books and volunteers, and often adapt an open model rather than train one from scratch.
Southeast Asia
AI Singapore's SEA-LION family covers about 11 languages of the region, including Indonesian, Thai, Vietnamese, Filipino, Tamil and Burmese. In December 2023 Singapore funded a two-year, S$70 million National Multimodal LLM Programme to extend it. Indonesia's Sahabat-AI and Thailand's Typhoon models adapt open base models such as Llama and Gemma to national languages.
Africa and beyond
Africa has over 2,000 languages, most with little digital text. The Masakhane research community builds datasets and models for African languages, and Johannesburg's Lelapa AI released InkubaLM in 2024, a small model for Swahili, Yoruba, isiXhosa, Hausa and isiZulu. BLOOM (2022) included about 20 Niger-Congo languages. In Latin America, a project led by Chile's national AI centre, CENIA, is building LatamGPT for the region.