Business

Mistral secures enterprise backing from ASML and Amadeus for data center buildout

Mistral secures commitments from ASML, Amadeus, and others to fund its gigawatt-scale data center buildout across Europe, positioning itself as a neocloud provider.

Mistral has secured multi-year commitments from ASML, Amadeus, and other enterprises to fund its gigawatt-scale data center buildout across Europe. The French AI firm is positioning itself as a neocloud, offering compute and cloud services. Mistral also opened its platform to third-party models and introduced regional inference controls.

Mistral, the French AI darling, has secured backing from a host of enterprise customers to support its ambitious data center buildout. The company is increasingly positioning itself as a European neocloud, offering compute and cloud services alongside its well-known AI models.

A coalition of enterprises

Dutch semiconductor equipment maker ASML, French logistics giant CMA-CGM, state-backed investor Caisse des Dépôts, consultancy firm Capgemini, and Spanish travel corporate Amadeus are among the anchor group of enterprises funding Mistral's AI infrastructure expansion across Europe.

These partners are making multi-year financial commitments that convert into "European Compute Units" (ECUs). ECUs give customers access to future compute and related services built by Mistral. The approach allows Mistral to fund its data centers with advance commitments, a model that the company says can "support infrastructure in Europe at a scale no participant could secure alone."

A gigawatt by 2030

Mistral plans to build one gigawatt (1,000 megawatts) of compute capacity by 2030, serving both its own needs and those of third-party customers. The company has already secured an $830 million loan to fund its first data center south of Paris, representing 44 megawatts of capacity.

CEO Arthur Mensch has estimated that one gigawatt of compute would represent a $50 billion investment. The company is unlikely to fund the project with VC backing alone, making enterprise commitments crucial.

A new kind of cloud provider

Mistral is expanding beyond its core model-building business to include cloud and compute services. The company is opening its platform to third-party open models, starting with GLM-5.2, an open-weights model developed by Chinese AI lab Z.ai. Customers will be able to choose other model providers for their AI workloads.

Mistral also introduced a feature that lets customers choose whether inference runs in the U.S. or Europe. This flexibility helps users align with data-residency, regulatory, and latency requirements. The move comes less than a month after Microsoft made a multi-billion-dollar commitment to using Mistral's compute infrastructure in Europe.

A European neocloud

If Mistral succeeds, it would become a sizable cloud player in Europe, comparable to Dutch neocloud Nebius, which recently contracted 750 MW of compute capacity in EMEA. The expansion signals a shift for Mistral from a model builder to a comprehensive infrastructure provider.

The enterprise backing is a vote of confidence in Mistral's vision. By turning its customers into investors of sorts, Mistral is securing both capital and committed demand. It's a model that could define the next phase of Europe's AI infrastructure.