Latam-Gpt: Free AI, Free Source and Latin American Cooperation

Latam-Gpt is new The big language model made in Latin America and for Latin America. The project, led by the Chilean Nonprofit Center of Chile (CENIA), aims to help the region achieve technological independence by providing an open source artificial intelligence model taught on Latin American languages ​​and texts.

“This is not just by a group or country in Latin America: this is a challenge that needs everyone to participate,” says Senate Manager Ilvaro Suto in an interview with Wired En Español. “Latam-GPT is a project that seeks to create a free, free, and most important artificial intelligence model. We have been working with a very low process for two years and bringing together citizens of different countries that want to cooperate. Recently, it is also seeing the initiatives from above, with governments starting their interest in this project.”

This project is due to its common spirit. “We do not seek to compete with Openai, Deepseek or Google. We want a special role model for Latin America and Caribbean, one of the cultural requirements and challenges that this requires, such as understanding different accents, region history and unique cultural aspects.”

Thanks to 33 strategic collaborations with institutions in Latin American and Caribbean, the project has collected millions of books as a collection of data over eight terabytes of text. This database has activated the development of a 50 billion parameter language model, a scale that compares it to Gpt-3.5, giving it a medium and high capacity for complex tasks such as reasoning, translation and associations.

Latam-GPT is trained in a regional database that collects information from 20 Latin America and Spain, and a total of 2,645,500 documents. Distribution of data shows significant concentrations in the largest countries in the region, with the Brazilian leader with 685,000 documents, followed by Mexico with 385,000, Spain with 325,000, Colombia with 220,000 and Argentina with 210,000 documents. These numbers represent the size of these markets, their digital development, and the availability of constructed content.

“At first, we will launch a language pattern. We expect its performance in general work close to large business models, but with excellent performance on specific issues for Latin America. The idea is that if we ask issues of our area, its knowledge will be much deeper.”

The first model is the starting point for family development of more advanced technologies in the future, including those with image and video and for scaling to larger models. “Since this is an open project, we want other institutions to use it. One group in Colombia can adapt it to the school education system, or one in Brazil can adapt it to the health sector. The idea is to open the doors of different organizations to produce specific models such as agriculture, culture and culture,” Cenia explains.