Latam-GPT is an artificial intelligence language model developed in Chile in collaboration with Brazil. It is designed to represent and understand the sociocultural context of Latin America and the Caribbean better than global models like ChatGPT.
Driven by the National Center for Artificial Intelligence (CENIA) and supported by regional institutions, Latam-GPT aims to be an ethical and responsible tool for people. Unlike other models, it is trained with specific data from the continent, allowing it to offer contextualized responses on Latin American topics.
The first free version of the model, scheduled to launch between June and July 2025, will function as a chatbot and understand Spanish, Portuguese, and English. Future phases include the inclusion of indigenous languages such as Mapudungun and Rapanui.
The project is supported by experts, organizations, and research centers from Mexico, Argentina, Colombia, Ecuador, the United States, Spain, Peru, and Uruguay.
Latam-GPT represents an advance towards the technological sovereignty of the region, allowing Latin American countries to develop local capabilities in artificial intelligence and reduce their dependence on foreign models.
The initiative also seeks to preserve and revitalize indigenous languages, incorporating translators for languages such as Rapanui, with plans to expand to other languages for use in virtual public service tools and personalized educational programs.
In summary, Latam-GPT is a collaborative project that seeks to create an artificial intelligence model adapted to the cultural and linguistic reality of Latin America and the Caribbean, promoting inclusion, diversity, and technological sovereignty in the region.