📊 Full opportunity report: OpenEuroLLM. The third path. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
OpenEuroLLM is a European consortium project to create open-source multilingual large language models. Despite progress, compute resource constraints remain a key challenge. The first models are expected by July 2026.
OpenEuroLLM, a European Union-funded consortium involving 20 organizations across academia, industry, and high-performance computing centers, is working toward creating open-source multilingual large language models. Despite achieving initial milestones, the project’s coordinator, Jan Hajič, has publicly stated that securing additional computing resources remains a significant challenge.
Launched in February 2025 with a €20.6 million EU grant from the Digital Europe Programme, OpenEuroLLM aims to develop multilingual LLMs accessible in the public domain. The project is led by Jan Hajič of Charles University in Prague and co-led by Peter Sarlin of Silo AI in Finland. Its consortium includes 20 partners, comprising universities, research institutes, AI companies, and supercomputing centers across Europe.
As of the March 6, 2026 progress report, the project has met initial goals but faces persistent hurdles related to computational capacity. Hajič emphasized that “significant challenges, especially in securing more compute for creating the final models, still remain,” highlighting that resource limitations are a shared obstacle across national and pan-European efforts.
Despite these constraints, the consortium continues to operate at scale, with the first models expected to be delivered by July 31, 2026. The project’s success hinges on overcoming the compute bottleneck, which is critical for training large, multilingual models at the desired quality and scope.
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
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.
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.
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.
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 Resource Constraints on European AI Progress
The ongoing compute limitations faced by OpenEuroLLM underscore a broader challenge for European AI development: scaling large language models within budget and infrastructure constraints. This bottleneck directly impacts the timeline and quality of the models, influencing Europe’s competitiveness in AI innovation. The project’s progress and eventual model deployment will serve as a benchmark for future European sovereign-LLM initiatives, shaping policy and investment decisions.
Microsoft's high-performance computing server with CD-ROM(Chinese Edition)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
European Sovereign-LLM Strategies and Resource Challenges
The European approach to developing sovereign language models has been characterized by three main strategies: Italy’s from-scratch investment in Minerva, Portugal’s continuation training with AMÁLIA, and the EU-wide pooled resources represented by OpenEuroLLM. Each approach reflects different assumptions about scale, institutional commitment, and resource sharing.
Previous essays by Thorsten Meyer have analyzed these models, revealing that all three are operating at a scale where resource constraints are evident. The recent progress report confirms that even the consortium-based OpenEuroLLM faces similar limitations, emphasizing that resource bottlenecks are a fundamental obstacle to achieving European autonomy in AI.
“Significant challenges, especially in securing more compute for creating the final models, still remain.”
— Jan Hajič
HPE NVIDIA Tesla V100 32GB HBM2 PCIe 3.0 x16 Passive GPU Computational Accelerator for AI Machine Learning HPC Deep Learning 699-2G500-0216-400 (Renewed)
- Architecture: NVIDIA Volta GV100 with CUDA and Tensor Cores
- Memory: 32GB HBM2 ECC with 900 GB/s bandwidth
- Interface: PCIe 3.0 x16 with 250W TDP
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unresolved Challenges in Computing Resources and Model Quality
It remains unclear how effectively the consortium will secure additional compute capacity before the July 2026 deadline. The extent to which resource limitations will impact the final model quality and scope is still uncertain, and whether new funding or infrastructure can alleviate these bottlenecks is yet to be determined.
LLM Systems Engineering: Training and Building Large Language Models – Engineering AI Models Through Fine-Tuning, Continued Pretraining, and From-Scratch Development
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Upcoming Model Release and Future Resource Strategies
The first models from OpenEuroLLM are scheduled for delivery by July 31, 2026. The project team plans to evaluate the models’ performance and scalability, which will influence future resource allocation and strategic decisions. Additional funding or infrastructure improvements could alter the current trajectory, but these remain uncertain as the project progresses.
ASUS Ascent GX10 AI Supercomputer, DGX Spark, NVIDIA GB10 Superchip, 128GB LPDDR5x, 1TB PCIe Gen4 NVMe SSD, Wi-Fi 7 & BT5.4, Agentic AI Ready, Supports OpenClaw, NemoClaw, Stackable Chassis
- AI Performance: Powered by NVIDIA GB10 Superchip with 1 petaFLOP
- High Memory Capacity: 128GB LPDDR5x memory for large models
- Fast Storage: 1TB PCIe Gen4 NVMe SSD
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is the main goal of OpenEuroLLM?
OpenEuroLLM aims to develop open-source, multilingual large language models for public use, representing a collaborative effort across Europe to foster AI sovereignty.
Why are compute resources a bottleneck?
Training large, multilingual models requires extensive computational power, which is limited by available hardware and funding, constraining the scale and quality of the models.
How does this project compare to national efforts like Minerva or AMÁLIA?
Unlike national projects that rely on local resources, OpenEuroLLM pools European infrastructure and expertise, but still faces the same fundamental resource constraints.
When will the first models be available?
The first models are expected to be delivered by July 31, 2026, with ongoing assessments of their performance and scalability planned afterward.
What are the implications if resource constraints persist?
Continued limitations could delay model deployment, reduce their scope, or impact their quality, affecting Europe’s competitiveness in AI development.
Source: ThorstenMeyerAI.com
