📊 Full opportunity report: ALIA. The Spanish answer. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Spain’s ALIA project, a 40-billion-parameter multilingual language model, has been publicly released. It is the largest publicly funded European AI initiative, emphasizing Spanish-language coverage and operational transparency, but benchmarks show it underperforms compared to Llama 2.
Spain has officially released ALIA, its largest publicly funded AI language model, trained on 9.37 trillion tokens across 35 European languages, including Spanish, marking a major milestone in European sovereign AI initiatives.
Developed under the auspices of the Barcelona Supercomputing Center (BSC-CNS) and coordinated by Spain’s Secretary of State for Digitalisation and Artificial Intelligence (SEDIA), ALIA is a 40-billion-parameter model trained from scratch with €90 million in MareNostrum 5 infrastructure upgrades and an additional €150 million dedicated to integration efforts. The model was released under the Apache License 2.0 on HuggingFace on April 22, 2025, and is designed to serve Spain’s strategic goal of fostering a multilingual, transparent, and publicly accessible AI infrastructure.
Operational benchmarks reveal that ALIA’s performance on standard NLP tasks like XNLI and SQuAD is below that of Llama 2, with accuracy rates of approximately 51.77% versus 66% on XNLI and 81.53% versus 93-94% on SQuAD, confirming a structural capability gap. The project emphasizes Spanish language coverage and multilingual support, aligning with its strategic positioning as a Position 3 project focused on operational relevance within the Spanish-speaking world, rather than competing for top performance globally. This approach may influence future European AI strategies, prioritizing widespread adoption and public infrastructure over solely performance metrics, and highlights the ongoing debate between Position 1 and Position 3 strategies.
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
AI Translator Language Translator Device Spanish & English Practice Companion – Supports 165 Languages，Tu compañero para practicar español e inglés,Travel Essentials traductor de idiomas portatil
1.Real-Time 165-Language Translation Instant two-waytranslation between Mexican Spanish & English with 0.5s lowlatency, perfect for restaurant, retail, hotel…
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.
AI Translator Language Translator Device Spanish & English Practice Companion – Supports 165 Languages，Tu compañero para practicar español e inglés,Travel Essentials traductor de idiomas portatil
1.Real-Time 165-Language Translation Instant two-waytranslation between Mexican Spanish & English with 0.5s lowlatency, perfect for restaurant, retail, hotel…
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 Networked Firearm Safety Protocol Frameworks: 2025 (Civil Blue Force Advanced Technology)
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.
Lark-2: Language Activity Resource Kit – Second Edition – Speech & Language Therapy Resource
Comprehensive Therapy Resource – Includes real objects, photographs, illustrations, and print materials to support a wide range of…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications of ALIA’s Strategic Positioning and Performance
ALIA’s development signifies Europe’s largest public investment in a national AI model, emphasizing transparency, multilingual support, and open-source availability. While its benchmark performance lags behind leading models like Llama 2, the project demonstrates Spain’s commitment to operational relevance within its linguistic and regional context. This approach may influence future European AI strategies, prioritizing widespread adoption and public infrastructure over solely performance metrics, and highlights the ongoing debate between Position 1 (global top-tier performance) and Position 3 (regional, operational, multilingual focus) strategies.
Spain’s Public AI Investment and Strategic Positioning
Spain’s ALIA project is part of a broader European effort to develop sovereign AI models, with previous initiatives including Portugal’s AMÁLIA, Italy’s Minerva, and the pan-European OpenEuroLLM. ALIA represents the most ambitious and publicly funded effort to date, with over €240 million in total public investment, including €90 million for hardware upgrades and €150 million for integration and deployment. The project operates within Spain’s national AI strategy, launched publicly in January 2025, and is led by the Barcelona Supercomputing Center under the coordination of SEDIA. The project’s emphasis on multilingual support and transparency aligns with European policies promoting digital sovereignty and AI independence.
“Our 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, ALIA project lead
Benchmark Performance and Strategic Ambiguity
While benchmarks confirm ALIA’s underperformance relative to Llama 2, it remains unclear how this gap will evolve with future training iterations or architectural improvements. Additionally, the strategic framing around Position 1 versus Position 3 remains contested, with some stakeholders emphasizing operational relevance over performance benchmarks. The long-term impact of ALIA’s open-source, multilingual approach on regional AI adoption and European sovereignty is still uncertain.
Next Steps for ALIA’s Deployment and Evaluation
Further benchmarking and real-world deployment will clarify ALIA’s operational capabilities and adoption levels within Spain and across Europe. The project team is expected to release updated versions and expand multilingual support, while policymakers will monitor its impact on regional AI sovereignty. Continued evaluation will determine whether ALIA can influence European AI sovereignty or if strategic adjustments are necessary to meet regional AI ambitions.
Key Questions
What is the main goal of Spain’s ALIA project?
The primary goal is to develop a publicly accessible, multilingual AI model that supports Spanish and co-official languages, focusing on operational relevance and regional adoption rather than global performance supremacy.
How does ALIA compare to other European AI initiatives?
ALIA is the largest publicly funded European national AI project in terms of investment and scale, surpassing efforts like Portugal’s AMÁLIA and Italy’s Minerva, with a focus on transparency and regional language support.
What are the benchmark performance results for ALIA?
ALIA’s benchmark scores on tasks like XNLI and SQuAD are below Llama 2, with approximately 51.77% versus 66% on XNLI and 81.53% versus 93-94% on SQuAD, indicating a structural performance gap.
What does ALIA’s development mean for European AI sovereignty?
It demonstrates Spain’s commitment to building a transparent, multilingual, open-source AI infrastructure, potentially influencing broader European policies on digital sovereignty and regional AI adoption strategies.
Source: ThorstenMeyerAI.com
