The OECD.AI Policy Navigator

Our policy navigator is a living repository from more than 80 jurisdictions and organisations. Use the filters to browse initiatives and find what you are looking for.

AMALIA – Automatic Multimodal Language Assistant with Artificial Intelligence


Added by:   OECD analyst
Added on:   17 Jul 2026
Updated by:   OECD analyst
Updated on:   17 Jul 2026

AMALIA is the first LLM designed for European Portuguese, with the aim of ensuring Portugal's linguistic and cultural sovereignty. The base version was launched in SEP25 and the full multimodal version (capable of interpreting text, images and video) arrives mid-2026. Use cases will focus on the areas of education, culture, speech, media, science and public administration. The model will be openly available so that it can be used and improved by scientists and development teams on any platform.

Name in original language

AMALIA – Assistente Multimodal Automático de Linguagem com Inteligência Artificial

Initiative overview

AMALIA is the first Large-Scale Language Model (LLM) designed specifically for European Portuguese language, using AI to process and generate natural language, with the ability to understand the linguistic diversity, cultural context, and historical references unique to Portugal. 

The main objective of AMALIA is to promote digital sovereignty, ensuring that sensitive national data can be processed securely and independently. It also aims to strengthen the national AI ecosystem by supporting innovation in key sectors such as education, culture, speech, media, science and public administration. 

AMALIA will be a multimodal model, capable of interpreting not only text but also speech, images and video, and its development implies an investment of €5,5 million, led by national research institutions and coordinated by the State Technological Reform Agency (ARTE). 

AMALIA shows strong performance in European Portuguese, outperforming other open models in lexicology and semantics. The base version is complete, and developed by over 60 researchers, reinforcing national digital sovereignty. Results are measured through technical benchmarking reports and evaluations in public services. Future impacts include broader use in education, public administration, and science, with the final version expected in mid-2026.

Key risks associated with AMALIA include establishing secure and compliant protocols for sharing non-public data, ensuring adequate funding and timely development of user interfaces for the defined use cases, and preventing data obsolescence. The critical conditions for the success of an initiative such as AMALIA include the effective productization of its use cases, ensuring that they move beyond prototypes into scalable, user-ready solutions, and the establishment of a robust inference infrastructure capable of supporting its widespread deployment across Public Administration.

All versions of the model, along with the training data, will be released as open source, making them freely available to researchers, public institutions, companies, and citizens. This initiative is the first major project under Portugal’s National Artificial Intelligence Agenda, reinforcing Portugal’s commitment to ethical, inclusive, and sovereign development. The current version is already available through the IAedu platform of the Foundation for Science and Technology (FCT), while the final version of AMALIA is expected to be launched in mid-2026.

About the policy initiative


Category:

  • AI policy initiatives, programmes and projects

Initiative type:

  • AI use cases/projects in the public sector

Status:

  • Proposed or under development

Start Year:

  • 2025

OECD AI Principles:


Other relevant urls: