The Government today announced that the model of Artificial Intelligence (IA) in Portuguese Amália will enter a new phase and have additional capabilities to “read, hear and see text, voice and images”.
© Shutterstock
04/09/2026
by
Lusa
IA Amalia
Nthe end of the Council of Ministers, the Minister of the Presidency, António Leitão Amaro said that it is planned to move from the prototype to the public service counters, since the idea is to “put Amália to help in meeting people”.
To train the model, access to data on equipment and public services such as “the National Library, the Cinema Library, archives and museums” is being considered, according to the minister.
Leitão Amaro also announced the intention to “start a dialogue” with media companies to “expand the possibility of an agreed access to their latest archives” by the AI model.
Portugal wants to be an example in the relationship between the national model of AI and media companies, respecting the concerns of companies regarding the “due compensation for the use of their archives”, he added.
The Amália – Automatic Multimodal Language Assistant with Artificial Intelligence is part of the National Agenda of AI, and was developed by a team composed by the Universidade Nova de Lisboa, Instituto Superior Técnico, Universidade de Coimbra, Universidade do Porto, Universidade do Minho and the Fundação para a Ciência e Tecnologia.
The process of creating the model began with the collection and processing of data in European Portuguese on a large scale, which were filtered based on its relevance and linguistic quality, and with knowledge collected by 2023.
To carry out the training of the models, large-scale computational infrastructure was used, through Iberian supercomputers, such as the Spanish Mare Nostrum 5, Portuguese Deucalion, and European, through the EuroHPC network.
A test version of the first Portuguese language model was completed in the first half of 2025, and the final version was presented on July 1.

