📊 Full opportunity report: ALIA. The Spanish answer. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Spain has launched ALIA, a 40-billion-parameter multilingual language model, with €240 million in public funding. While operationally below Llama 2 benchmarks, it aims to serve the Spanish-speaking world, highlighting strategic positioning between European sovereignty and operational performance.
Spain’s government has officially launched ALIA, a 40-billion-parameter multilingual language model trained with €240 million in public funds, marking Europe’s largest publicly funded national AI project. The initiative aims to enhance Spanish-language AI capabilities and promote European sovereignty in artificial intelligence, with operational benchmarks indicating it is below some leading models like Llama 2 but strategically focused on Spanish and co-official languages.
Developed under the auspices of the Barcelona Supercomputing Center (BSC-CNS) and led by the Secretary of State for Digitalisation and Artificial Intelligence (SEDIA), ALIA-40B was trained on 9.37 trillion tokens across 35 European languages and 92 programming languages. The project received €90 million for MareNostrum 5 upgrades and €150 million for ALIA integration into industry, totaling over €240 million in public funding.
The model was released under the Apache License 2.0 on HuggingFace in April 2025 and is designed to serve the Spanish-speaking world, emphasizing multilingual coverage with an oversampling of Spanish. Benchmark results show ALIA-40B’s performance at 51.77% on XNLI in English and 81.53% on SQuAD in English, which are below Llama 2’s scores of approximately 66% and 93-94%, respectively. This confirms a structural capability gap compared to some leading models at similar scales.
Despite its sub-Llama-2 operational performance, project leaders, including Josep M. Martorell, emphasize that the goal is to maximize adoption within the Spanish-speaking community, aligning with a Position 3 strategic profile focused on multilingual specialization rather than global performance dominance.
ALIA.
The Spanish
answer.
€240M+ Spanish public funding · ALIA-40B + Salamandra family · 9.37T tokens · 35 European languages + 92 programming languages · MareNostrum 5 · Apache 2.0 release. The largest publicly funded European national-AI project by cumulative scope — and the empirical test case for the Position 1 vs Position 3 strategic-positioning argument.
This is the tenth standalone essay in the European sovereign-LLM track and the third Tier 2 expansion piece. ALIA is Spain’s institutional answer — the largest EU member state by GDP not yet documented in the track. The project markets itself as Position 1 + Position 2 simultaneously — “Europe’s first public multilingual foundational model.” The benchmark evidence (ALIA-40B 51.77% XNLI_en vs Llama 2 66%) confirms the structural capability gap from Finding 1 of the synthesis essay. The Position 3 framing — Martorell’s “most widely adopted in the Spanish-speaking world” — is operationally honest. €90M MareNostrum 5 upgrade + €150M company integration = €240M+ cumulative scope. Apache 2.0 open-source release + AESIA validation + co-official languages oversampling. Both can be true at once. The Spanish public discourse would benefit from explicit Position 3 strategic positioning.
Six models. Apache 2.0.
The ALIA family operates as a tiered model portfolio. ALIA-40B is the flagship at 40 billion parameters; the Salamandra family scales down to 7B, 2B and instruct-tuned variants; mRoBERTa provides the foundational multilingual baseline. All released under Apache License 2.0 on April 22, 2025 at the HispanIA 2040 event — “Public Code, Public Money” approach.
multilingual
MN5 LLM
edge
target
instruct
encoder
multilingual AI language model
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Four official. Oversampled by factor of 2.
ALIA’s distinctive multilingual coverage strategy. The four co-official Spanish languages are oversampled by factor of 2 in the training corpus — structurally distinct from Apertus’s broad 1,811-language coverage approach. The strategy targets deep coverage of Spanish co-official languages rather than maximum language breadth.
Spanish language AI tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
ALIA-40B vs Llama 2. 14-point gap.
The empirical evidence Finding 1 of the synthesis essay needed. ALIA-40B at 40 billion parameters with €240M+ public funding and 8+ months MareNostrum 5 training achieves performance below Llama 2 — a 2023 frontier model released approximately 18 months before ALIA-40B. The capability gap is real and consistent with six of seven prior national-project answers documented in the track.
AI development software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Two pilots. Public administration deployment.
The operational deployment targets that validate the Position 3 + Position 4 framing. Public administration deployment is the structurally credible Position 3 + Position 4 strategic positioning — captive demand from Spanish public institutions where Spanish-language specialization is operationally distinctive.
The work is real across the Spanish ALIA case. €240M+ public funding committed. 40B parameter from-scratch model trained on 9.37 trillion tokens. Salamandra family released under Apache 2.0. AESIA validation aligned with EU AI Act transparency standards. Two pilot applications shipped — Tax Agency chatbot and primary care medicine heart failure diagnosis. The Position 1 framing is operationally misleading. ALIA-40B performance below Llama 2 confirms the structural capability gap. The Position 3 framing is operationally honest — Spanish-speaking world adoption, co-official languages oversampling, public administration deployment. Both can be true at once. The Spanish public discourse would benefit from explicit Position 3 strategic positioning.
European sovereign AI platform
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications of ALIA’s Strategic Positioning for European AI
ALIA exemplifies Spain’s commitment to developing a sovereign AI infrastructure tailored to the Spanish-speaking world, emphasizing multilingual capabilities and transparency. While operational benchmarks indicate it is not yet competitive with top models like Llama 2, its strategic focus on regional adoption and co-official languages underscores a broader European debate about balancing performance with sovereignty and language representation. This project sets a precedent for other EU nations pursuing similar national AI initiatives amid ongoing discussions about AI regulation, data sovereignty, and language inclusion.Background of Spain’s AI Sovereignty Efforts and ALIA Development
Spain’s ALIA project is part of a broader national AI strategy initiated in 2019, with public investments totaling over €240 million. It follows previous efforts such as the Language Technologies Plan and projects like AINA and ILENIA, which aimed to develop regional language AI tools. The project is coordinated by the Barcelona Supercomputing Center, leveraging MareNostrum 5’s 4,480 NVIDIA H100 GPUs, and aligns with Spain’s goal to establish a sovereign AI infrastructure capable of supporting government, industry, and societal needs.
Compared to other European national models like Portugal’s AMÁLIA, Italy’s Minerva, and the pan-European OpenEuroLLM, ALIA represents the most ambitious publicly funded effort in terms of scale and scope. It is also a structural response to the strategic debate within Europe about whether to prioritize global performance or regional linguistic and sovereignty considerations in AI development.
“The goal is not to be the best-performing LLM in the world, but the most widely adopted in the Spanish-speaking world.”
— Josep M. Martorell
Operational Performance Versus Strategic Goals Unclear
While benchmark results confirm ALIA-40B’s performance is below top models like Llama 2, it remains uncertain how this gap will impact real-world adoption and industry integration in Spain and wider Europe. The long-term effectiveness of the model in practical applications and its ability to compete globally are still developing issues. Additionally, the strategic implications of prioritizing regional language coverage over raw performance are subject to ongoing debate within the AI community and policymakers.
Next Steps for ALIA Deployment and Performance Evaluation
Further testing and deployment of ALIA in industry and government sectors are expected over the coming months. The project will likely undergo ongoing benchmarking against emerging models, with adjustments aimed at improving performance while maintaining its regional and multilingual focus. Additionally, discussions around regulatory standards, transparency, and language inclusion will shape the model’s evolution and integration into Spain’s digital infrastructure.
Key Questions
What is the main goal of Spain’s ALIA project?
ALIA aims to develop a multilingual language model focused on Spanish and co-official languages, emphasizing regional adoption and sovereignty rather than global performance supremacy.
How does ALIA compare to other European AI models?
Operationally, ALIA-40B currently performs below models like Llama 2, but it is the largest publicly funded European national AI project, with a strategic focus on regional language coverage and transparency.
What are the main challenges facing ALIA?
The primary challenge is its benchmark performance being below leading models, which may affect competitiveness. Its success will depend on regional adoption and integration into Spanish industry and government.
Will ALIA be open source?
Yes, ALIA-40B was released under the Apache License 2.0 on HuggingFace in April 2025, supporting transparency and wider adoption.
What is the strategic significance of ALIA for Europe?
ALIA demonstrates Europe’s approach to balancing AI sovereignty, regional language support, and transparency, serving as a model for other EU nations pursuing similar initiatives.
Source: ThorstenMeyerAI.com
