📊 Full opportunity report: OpenEuroLLM. The third path. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
OpenEuroLLM, a major European AI project involving 20 organizations and funded by €20.6M from the EU, is struggling with compute resource constraints. This reveals the structural limits of pan-European AI initiatives and their reliance on shared supercomputing resources.
OpenEuroLLM, the pan-European consortium aiming to create an open-source multilingual large language model, has publicly acknowledged that securing additional computational resources remains a major challenge.
Launched in February 2025 and currently one year into a three-year development cycle, OpenEuroLLM is coordinated by Jan Hajič at Charles University in Prague and co-led by Peter Sarlin of Silo AI in Finland. Funded by €20.6 million from the EU’s Digital Europe Programme as part of a total €37.4 million budget, the project involves 20 organizations across academia, industry, and high-performance computing centers across Europe.
According to Hajič’s March 6, 2026 progress report, despite achieving initial goals, the consortium faces significant hurdles in securing enough compute power to develop the final models. He emphasized that even at a pan-European scale, computational capacity remains a bottleneck, constraining progress and limiting the scope of model training efforts.
This challenge reflects a broader pattern seen in European sovereign AI initiatives, where resource constraints—particularly compute—limit the scale and potential impact of models developed under different strategic approaches. The project’s first models are expected in July 2026, and their performance and scale will be critical indicators of the consortium’s success and the structural limits of such large-scale collaborative efforts.
OpenEuroLLM.
The third
path.
€37.4M EU budget, 20 organizations, four major EuroHPC supercomputers, 35 target languages. And the project’s coordinator says: “significant challenges in securing more compute still remain.”
Italy bet national. Portugal bet continuation. The EU bet consortium. OpenEuroLLM — coordinated by Jan Hajič at Charles University Prague, co-led by Peter Sarlin at AMD-owned Silo AI — is what the pan-European pooled-resources answer looks like in operational form. And the project lead is publicly stating that even at pan-European pooled scale, compute is the bottleneck. Each of the three sovereign-LLM answers, examined honestly, surfaces a complication the press coverage downplays.
Even at pan-European scale, compute is the bottleneck.
From the OpenEuroLLM first-year progress report, March 6, 2026. The single most important sentence in the public documentation of the project. The pan-European consortium answer — explicitly designed as the response to individual national projects’ resource constraints — is itself constrained by the same resource that limits national projects.
First-year progress and next steps · March 6, 2026
Performance Tuning with SQL Server Dynamic Management Views (High Performance SQL Server)
Used Book in Good Condition
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
12 universities. 6 companies. 3 HPC centers. One conspicuous absence.
The OpenEuroLLM consortium combines academic NLP research, commercial AI capability, and EuroHPC supercomputing infrastructure across multiple European nations. The breadth is the strategic bet. The breadth is also the operational complication.
multilingual AI training hardware
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Eleven deliverables. Two shipped. Nine pending.
From the official deliverables roadmap. As of mid-May 2026, only two of eleven deliverables have shipped — both from July 2025. The July 31, 2026 cluster — first models, initial dataset, evaluation code — is when OpenEuroLLM becomes empirically comparable to Minerva and AMÁLIA.
AI at Scale: Building Supercomputing Platforms for Large AI Models
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Three answers. Three structural findings.
The Minerva from-scratch path. The AMÁLIA continuation path. The OpenEuroLLM consortium path. Each project surfaces an empirical complication the press coverage downplays. Each finding is harder than the framing it’s wrapped in.
Three projects. Three findings. Each one harder than the framing it’s wrapped in. Each answer is valid for its specific positioning and resource context. None of the three is “the right answer” in the abstract. The strategic discourse benefits from treating all three as data points in the same empirical experiment.
Mastering Large Language Models with NVIDIA RTX 3090: A Comprehensive Guide
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
First models in six weeks. Three scenarios.
The July 31, 2026 first-models deliverable is the strategic moment for OpenEuroLLM specifically and for the European sovereign-LLM movement broadly. Three scenarios are plausible. The structurally honest framing will require acknowledging whatever the empirical results actually show.
OpenEuroLLM is one valid answer to the European sovereign-LLM question. AMÁLIA is another. Minerva is a third. Mistral is potentially a fourth — the commercial-frontier answer this essay track examines next. The strategic discourse benefits from treating all of them as complementary experiments in the same empirical question. More analysis like this is needed. Not less.
Implications of Compute Constraints for European Sovereign AI
The acknowledgment of resource limitations by OpenEuroLLM underscores a fundamental challenge for European AI sovereignty: the reliance on shared supercomputing infrastructure may inherently cap the scale and ambition of publicly funded models. This raises questions about the long-term viability of pan-European collaborations versus national or private efforts in developing competitive AI capabilities.
Furthermore, the project’s struggles highlight the importance of securing dedicated, scalable compute resources to realize the full potential of multilingual, open-source LLMs. As models grow in size and complexity, resource constraints could slow progress, impacting Europe’s ability to develop independent, high-performance AI systems that meet strategic and economic needs.
European Sovereign-LLM Strategies and Resource Challenges
The European approach to sovereign LLM development has been characterized by three main strategies: Italy’s from-scratch investment in models like Minerva, Portugal’s continuation training approach exemplified by AMÁLIA, and the pan-European consortium model represented by OpenEuroLLM. Each approach reflects different assumptions about investment scale, institutional commitment, and resource sharing.
Previous essays by Thorsten Meyer have examined these models, noting that all face significant resource constraints. Minerva’s development was limited by access to supercomputing resources, while Portugal’s AMÁLIA focused on continuation training within existing infrastructure. The current state of OpenEuroLLM, with its explicit resource bottleneck, confirms that these challenges persist at a continental level. The consortium’s progress and upcoming model releases will serve as critical benchmarks for evaluating the effectiveness of pooled resources versus national investments in Europe’s AI landscape.
“Even at pan-European pooled scale, compute is the bottleneck.”
— Jan Hajič, Charles University
Unresolved Questions About Model Performance and Scalability
It remains unclear how significantly the compute limitations will impact the quality, size, and multilingual capabilities of the models ultimately produced by OpenEuroLLM. The first models are scheduled for July 2026, but their effectiveness and scope are still uncertain, and it is not yet clear whether additional resources will be secured in time.
Furthermore, the broader implications for Europe’s AI sovereignty strategy, including potential shifts toward more national-focused efforts or increased private sector involvement, are still developing and depend on the project’s upcoming results.
Upcoming Model Releases and Resource Allocation Decisions
The next critical milestone for OpenEuroLLM is the release of its first models in July 2026, which will provide concrete data on the impact of current resource constraints. The project team is expected to evaluate whether additional compute resources can be secured before this deadline or if adjustments to scope and ambition are necessary.
Additionally, the broader European AI community will closely monitor the project’s outcomes to inform future strategies, including potential expansion of dedicated supercomputing infrastructure or shifts in collaborative models.
Key Questions
What is the main goal of OpenEuroLLM?
OpenEuroLLM aims to develop an open-source, multilingual large language model for European languages, fostering sovereignty and collaboration across the continent.
Why are compute resources a major challenge?
Training large language models requires significant computational power, which is limited by available supercomputing infrastructure across Europe. Securing enough compute remains a key bottleneck for progress.
How does this project compare to national efforts like Minerva or AMÁLIA?
While Minerva and AMÁLIA focus on from-scratch or continuation training at national levels, OpenEuroLLM attempts a pooled, pan-European approach, which faces additional resource and coordination challenges.
Will the upcoming models meet European language needs?
It is still uncertain; the models’ quality and multilingual capabilities will depend heavily on the available compute resources and training scope, which are currently limited.
What are the implications for Europe’s AI sovereignty?
The resource constraints highlight the importance of developing dedicated infrastructure and strategic investments to ensure Europe can develop competitive and independent AI systems.
Source: ThorstenMeyerAI.com
