📊 Full opportunity report: ALIA. The Spanish answer. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Spain’s ALIA-40B, a €240M public-funded multilingual large language model, has been released. It emphasizes Spanish-language coverage and operational transparency, but benchmarks show it lags behind Llama 2. The project reflects Spain’s strategic focus on widespread adoption over top performance.
Spain has officially launched ALIA-40B, a 40-billion-parameter multilingual large language model (LLM), developed with over €240 million in public funding. The model, trained on 9.37 trillion tokens across 35 European languages and 92 programming languages, aims to serve as Spain’s institutional answer to European sovereignty and AI independence. The release, coordinated by the Barcelona Supercomputing Center and led by the Secretary of State for Digitalisation and Artificial Intelligence, marks the country’s most ambitious public AI project to date. For more on AI investment trends, see this analysis of hyperscaler CapEx.
ALIA-40B was trained on MareNostrum 5’s 4,480 NVIDIA H100 GPU partition, utilizing €90 million for infrastructure upgrades and €150 million for project integration, making it the largest publicly funded European national AI effort in scope. The model was released under the Apache License 2.0 on HuggingFace on April 22, 2025, and has been validated by AESIA for transparency and co-official language coverage. Despite its extensive multilingual scope, benchmark results indicate it underperforms compared to Llama 2, with 51.77% accuracy on XNLI in English versus Llama 2’s 66%, and 81.53% on SQuAD in English versus Llama 2’s approximately 93-94%. The project’s leadership emphasizes widespread adoption within the Spanish-speaking world rather than top benchmark performance, positioning ALIA as a strategic, operationally credible initiative rather than a global top-tier model.
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.
large language model for developers
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 translation 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 AI research tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications of ALIA-40B for European AI Sovereignty
Spain’s ALIA-40B demonstrates a significant investment in public AI infrastructure, emphasizing multilingual coverage and transparency aligned with European sovereignty goals. While benchmark performance lags behind leading models like Llama 2, the project’s focus on widespread adoption and operational credibility reflects a strategic shift in national AI efforts. This approach may influence future European AI initiatives by prioritizing language coverage, transparency, and public-sector integration over pure performance, impacting the broader landscape of AI development and deployment across the continent.
Spain’s Public AI Investments and Strategic Positioning
Spain’s ALIA project is part of a broader trend of European nations investing heavily in sovereign AI initiatives. Learn more about European AI sovereignty efforts here. The project is the tenth standalone effort documented in the European sovereign-LLM track, following initiatives from Portugal, Italy, France, Germany, Switzerland, and pan-European consortia. Unlike some efforts focused on high-performance benchmarks, ALIA’s strategy emphasizes multilingual coverage, transparency, and integration into public and private sectors. The project’s €240 million public funding surpasses previous national efforts, reflecting Spain’s commitment to establishing an operationally credible, publicly accessible AI infrastructure that aligns with European sovereignty and digital independence goals.
“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 Capabilities and Benchmark Performance Gaps
While ALIA-40B has been officially released and validated for transparency, its benchmark performance remains below that of leading models like Llama 2, raising questions about its competitiveness in high-stakes AI applications. It is not yet clear how the model will perform in real-world deployment scenarios or how it will be adopted across sectors. Additionally, the long-term impact of its multilingual coverage and the extent of its integration into Spanish industry and government are still developing.
Future Adoption, Benchmarking, and Strategic Adjustments
Next steps include monitoring the model’s adoption within Spain’s public sector and industry, evaluating its performance in practical applications, and assessing user feedback. This article discusses strategic AI investments that could influence future adoption. Further benchmarking and comparative studies are expected to clarify its operational strengths and weaknesses. Spain’s government and project leaders may also refine the model’s capabilities or develop complementary tools to enhance its competitiveness, while continuing to emphasize transparency and multilingual coverage as core strategic pillars.
Key Questions
What is ALIA-40B?
ALIA-40B is a 40-billion-parameter multilingual large language model developed by Spain’s public AI initiative, trained on extensive European language data, and released under open-source license.
Why is ALIA-40B significant for Europe?
It represents Europe’s largest public investment in a sovereign AI model, emphasizing multilingual coverage, transparency, and strategic independence from non-European models.
How does ALIA-40B compare to other models like Llama 2?
Benchmark results show ALIA-40B underperforms compared to Llama 2 in key tasks, indicating a focus on operational transparency and language coverage over top benchmark performance.
What are the strategic goals behind ALIA?
The project aims to foster widespread adoption of Spanish and European languages, ensure transparency, and strengthen Spain’s digital sovereignty, rather than competing solely on benchmark metrics.
What are the next steps for ALIA?
Future developments include evaluating real-world performance, expanding adoption, and possibly refining the model to improve benchmarks while maintaining its strategic focus on language coverage and transparency.
Source: ThorstenMeyerAI.com
