📊 Full opportunity report: ALIA. The Spanish answer. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Spain’s ALIA project, with €240 million in public funding, has released the ALIA-40B multilingual model trained on 9.37 trillion tokens. While claiming to be Europe’s first public multilingual LLM, benchmark results show it underperforms compared to Llama 2, highlighting a structural capability gap.
Spain has publicly released ALIA-40B, a multilingual large language model (LLM) trained on over 9.37 trillion tokens across 35 European languages, marking the country’s largest publicly funded AI initiative to date.
The project, led by the Barcelona Supercomputing Center (BSC-CNS) and coordinated by the Secretary of State for Digitalisation and Artificial Intelligence (SEDIA), involves a €240 million investment from the Spanish government, including €90 million for MareNostrum 5 upgrades and €150 million for ALIA integration into industry.
ALIA-40B is trained from scratch on a massive dataset and released under the Apache License 2.0 on HuggingFace on April 22, 2025. The model is designed to serve as Spain’s institutional answer to European sovereignty questions in AI, emphasizing multilingual capabilities with a focus on Spanish and co-official languages.
Benchmark results show ALIA-40B’s performance against Llama 2 at 40B parameters, with significantly lower scores—51.77% on XNLI_en versus Llama 2’s 66%, and 81.53% on SQuAD_en versus Llama 2’s 93-94%. These results confirm a structural capability gap, indicating that the project’s operational performance is below that of leading models.
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 Translation Mastery: Building Accurate, Culturally Sensitive Language Tools and Global Communication Systems in 2026
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.
Build a Large Language Model (From Scratch)
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.
GPT AI Chat: ChatGPT Chatbot
GPT-4 powered
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.
Hugging Face Transformers Cookbook: Fine-Tuning Open-Source Models for Niche Industry Tasks
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications of ALIA-40B’s Performance and Strategic Positioning
While ALIA-40B marks a milestone as Europe’s largest publicly funded national AI project, its benchmark results reveal a capability gap compared to commercial models like Llama 2. The project’s framing emphasizes multilingual and co-official language coverage, aligning with Spain’s strategic aim to foster widespread adoption in the Spanish-speaking world.
However, the benchmark data suggests that, operationally, ALIA-40B is not yet competitive at the highest performance levels, confirming the project’s positioning as a Position 3, focused on regional and linguistic relevance rather than pushing the performance envelope. This distinction influences how the project might shape Spain’s AI policy and its role within European sovereignty efforts.
Spain’s National AI Strategy and European Sovereignty Efforts
Spain’s ALIA project is part of a broader national AI strategy launched in early 2025, with €240 million in public funding dedicated to developing a sovereign AI infrastructure. It follows previous European and national projects such as Portugal’s AMÁLIA, Italy’s Minerva, and the pan-European OpenEuroLLM, all of which aimed to develop large language models with varying scopes and funding levels.
Unlike some European projects that focus on commercial applications or enterprise sovereignty, ALIA emphasizes multilingual capabilities, open-source deployment, and alignment with Spanish government priorities. The project’s strategic framing aligns with a Position 3 approach—focused on regional adoption and linguistic relevance—despite claims of being a Position 1 model aimed at global competitiveness.
“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 Claims Under Review
While benchmark results clearly show ALIA-40B’s lower performance compared to Llama 2, it remains unclear how the model’s operational effectiveness will translate into real-world applications or adoption within Spain and Europe. The strategic framing as a Position 1 attempt contrasts with the empirical data supporting a Position 3 profile, creating ongoing debate about the project’s future trajectory and impact.
Next Steps for ALIA and European Sovereign AI Strategies
Further benchmarking and testing are expected as the project moves into industry deployment phases, with ongoing assessments of performance and adoption. The Spanish government and project leaders may also refine their messaging to better align strategic claims with operational realities, potentially influencing future funding and policy decisions. European AI policymakers will closely monitor ALIA’s development as a case study for national sovereignty and multilingual AI deployment.
Key Questions
What is ALIA-40B and why is it significant?
ALIA-40B is Spain’s publicly funded multilingual language model trained on over 9.37 trillion tokens. It is significant as Europe’s largest public AI project, aiming to foster regional adoption and sovereignty, though benchmark results show it underperforms compared to commercial models like Llama 2.
How does ALIA compare to other European AI projects?
Unlike some projects focused on enterprise sovereignty or pan-European models, ALIA emphasizes multilingual capabilities and regional relevance, with a strategic framing as a Position 3 model, despite claims of being a Position 1 project.
What are the main limitations of ALIA-40B based on current data?
Benchmark results indicate that ALIA-40B’s performance is below that of leading models like Llama 2, highlighting a structural capability gap. Its operational effectiveness in real-world applications remains to be seen.
What is the future outlook for ALIA and Spain’s AI strategy?
Next steps include further testing, deployment, and strategic messaging adjustments. The project will be monitored for its adoption and impact within Spain and Europe, shaping future sovereignty policies.
Source: ThorstenMeyerAI.com
