Portugal launches its own open-source AI model, “Amália”
Amália comes as an alternative Portuguese-language large language model (LLM) that will be released under an open-source license.
Image by Cybernews
Amália comes as an alternative Portuguese-language large language model (LLM) that will be released under an open-source license.
- Portugal is launching the first large language model designed specifically for European Portuguese.
- It will serve as an open technological platform that enables public institutions and enterprises to develop their own AI applications.
- An increasing number of European countries, including Switzerland and Poland, are introducing their domestic LLMs.
The government-backed Amália, which is the first LLM designed specifically for European Portuguese, will be officially launched on July 1st, 2026, according to Portugal Resident.
The model, its dataset, and source code will all be released under an open-source license.
Amália will serve as an open technological platform that public institutions, enterprises, and researchers can build upon to create their own AI applications.
The LLM won’t be launched as a consumer-facing chatbot, such as ChatGPT or Gemini.
The name Amália is short for Assistente Multimodal Automático de Linguagem com Inteligência Artificial (Automatic Multimodal Language Assistant with Artificial Intelligence), and refers to Amália Rodrigues, iconic singer of the traditional Portuguese genre fado.
The LLM was built upon the existing open-source European model, EuroLLM-9B, and then expanded with additional European Portuguese training data.
It will first be deployed in the country’s public sector, targeting education, defense, and healthcare services, among others.
Have thoughts about this topic? Others do, too. Join them in the discussion.
While most frontier LLMs originate from the US and China, with the exception of France’s Mistral, European countries have been developing their own national solutions.
Switzerland unveiled its first national LLM, Apertus, in 2025, for research, education, and commercial applications. The model, its architecture, and training data are openly accessible.
Poland’s domestic LLM, called Polish Large Language Universal Model (PLLuM), was launched in early 2025. The creators then said PLLuM was “tailored to the specificities of the Polish language” and will support the public services system.
The need for local AI capabilities is especially evident after the Donald Trump administration ordered the US company Anthropic to suspend access to its most advanced models, Fable 5 and Mythos 5, to non-American citizens.
Unlock more exclusive Cybernews content on YouTube.