📊 Full opportunity report: OpenEuroLLM. The third path. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
OpenEuroLLM, a pan-European AI project funded by €20.6M from the EU, is progressing but still faces significant compute resource challenges. Its first models are due July 2026, and the project exemplifies Europe’s collaborative approach to sovereign AI.
European AI researchers and institutions, led by the OpenEuroLLM project, report that despite progress, significant challenges remain in securing enough computing power to develop their multilingual large language models (LLMs). This highlights ongoing resource constraints facing Europe’s sovereign AI ambitions.
OpenEuroLLM is a €37.4 million project funded primarily by the EU’s Digital Europe Programme, involving 20 organizations across universities, industry, and high-performance computing centers. Coordinated by Jan Hajič at Charles University and co-led by Peter Sarlin at Silo AI, the initiative aims to create open-source multilingual LLMs for public and research use.
As of the March 6, 2026 progress report, the project has achieved its first-year goals but faces persistent challenges in securing additional compute capacity needed for training the final models. According to Hajič, “significant challenges, especially in securing more compute for creating the final models, still remain.” The first models are scheduled for release by July 31, 2026, but resource limitations could impact this timeline.
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
HKUXZR C612 NAS Motherboard LGA2011-3, 10x SATA 6Gbps, 4X 2.5GbE Intel i226-V, 2X M.2 NVMe, 2X PCIe x16, DDR4, Server Workstation ITX Mainboard for Xeon E5 V3/V4 24 * 24cm
【High Performance Processor Support】 Supports Intel Xeon E5-V3/V4 series processors (LGA2011-3 socket), as well as Core i7/i9 series…
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 large language model 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.
EU supercomputers for AI research
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.
Learn Google Cloud Platform: Build Scalable Apps with GCP Compute, Kubernetes, Cloud Run, and BigQuery.
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 Bottlenecks for European Sovereign AI
This development underscores the fundamental resource constraint facing European AI initiatives: compute power. Despite collaborative efforts and substantial funding, the inability to scale compute resources may limit the quality, scope, and timeline of Europe’s sovereign LLMs, affecting strategic independence in AI technology.
It also highlights that even large, well-funded pan-European projects face structural limitations similar to national efforts, raising questions about the most effective models for developing competitive AI at scale within Europe.
European Sovereign-LLM Strategies and Resource Challenges
Europe’s approach to sovereign AI development has been characterized by multiple strategies, including Portugal’s continuation pre-training (AMÁLIA), Italy’s from-scratch investment (Minerva), and the collaborative consortium model exemplified by OpenEuroLLM. Each approach reflects different assumptions about investment scale, architectural commitment, and institutional coordination.
Previous efforts, such as Portugal’s AMÁLIA and Italy’s Minerva, have demonstrated that resource limitations—particularly in compute—are a persistent barrier. The OpenEuroLLM project, launched in early 2025, is the latest attempt to pool resources across multiple countries, but its progress indicates that resource constraints remain a key challenge in realizing fully functional, large-scale multilingual models.
“”Significant challenges, especially in securing more compute for creating the final models, still remain.””
— Jan Hajič
Unresolved Impact of Compute Limitations on Model Quality
It is not yet clear how significantly the compute bottleneck will affect the quality and capabilities of the first models scheduled for release in July 2026. The final models’ performance and the project’s overall success remain uncertain until those models are completed and evaluated.
Next Milestones and Model Release Expectations
The project aims to deliver its first models by July 31, 2026. The upcoming months will determine whether additional compute resources can be secured and whether the models meet the project’s performance and multilinguality goals. The first models’ release will serve as a key indicator of the project’s trajectory and Europe’s capacity to develop sovereign AI at scale.
Key Questions
What is the main goal of the OpenEuroLLM project?
The project aims to develop open-source, multilingual large language models for public and research use across Europe, fostering sovereign AI capabilities.
Why is compute power a bottleneck for OpenEuroLLM?
Training large multilingual models requires immense computational resources. Despite pooling efforts, securing enough compute remains a challenge, impacting model development timelines and quality.
How does OpenEuroLLM compare to national projects like Minerva and AMÁLIA?
OpenEuroLLM adopts a consortium approach pooling resources across countries, whereas Minerva and AMÁLIA are more nationally focused. All face similar resource constraints, but OpenEuroLLM’s scale makes the bottleneck more prominent.
What are the risks if compute resources are insufficient?
Insufficient compute could delay model release, reduce model quality, or limit multilingual capabilities, impacting Europe’s strategic independence in AI technology.
Source: ThorstenMeyerAI.com
