Latam-GPT — Latin America's First Sovereign Foundation Model
Chile launched Latam-GPT, the first open-source AI language model designed for Latin America. 70 billion parameters built on Meta's Llama 3.1 architecture, trained in Spanish and Portuguese with plans for indigenous languages. Coordinated by CENIA (National Center of Artificial Intelligence) with 60+ institutions and 200 specialists across 8 countries. Funded with $550K from CENIA + Development Bank of Latin America (CAF). A $4.5M supercomputer at the University of Tarapacá in northern Chile expected operational H1 2026 for subsequent versions. The model is positioned as public infrastructure — open for governments, universities, startups, and companies to build upon.