
Image by Christian Rucinski, from Unsplash
Latam-GPT: Latin America Develops Open Source AI
A new artificial intelligence project named Latam-GPT is being developed in Latin America to foster technological independence for the region.
In a rush? Here are the quick facts:
- CENIA leads the project with 33 regional institutional partners.
- The model uses over 2.6 million documents from 20 countries.
- Latam-GPT has 50 billion parameters, comparable to GPT-3.5.
The Chilean National Center for Artificial Intelligence (CENIA) is leading this initiative, and received support from 33 institutions throughout the region, as detailed in a report by WIRED.
“This work cannot be undertaken by just one group or one country in Latin America: It is a challenge that requires everyone’s participation,” Álvaro Soto, director of CENIA said to WIRED. “Latam-GPT is a project that seeks to create an open, free, and, above all, collaborative AI model,” he added.
The model was trained on more than 2.6 million documents from 20 countries, with Brazil, Mexico, Colombia, Argentina, and Spain contributing the most data. The system contains 50 billion parameters, comparable to GPT-3.5, and can perform translation, reasoning, and culturally relevant tasks.
“We’re not looking to compete with OpenAI, DeepSeek, or Google,” Soto said to WIRED. “We want a model specific to Latin America and the Caribbean, aware of the cultural requirements and challenges that this entails,” he added.
A $10 million supercomputing center at the University of Tarapacá in Chile provides the infrastructure for training the model, using state-of-the-art NVIDIA GPUs.
The first version of Latam-GPT will become available this year, followed by plans to develop image and video AI capabilities and model adaptations for education, healthcare, and other sectors.
Soto emphasizes the regional focus: “Success would mean that Latam-GPT has played an important role in the development of artificial intelligence in this region.”
The project is particularly significant given the growing global AI divide, where most advanced AI data centers are concentrated in the U.S., China, and the EU.
By creating regional computing capacity and culturally relevant AI, Latam-GPT reduces dependency on foreign technology, helps retain skilled professionals, and allows local solutions tailored to Latin American challenges.