PARIS AND VELDHOVEN, September 9, 2025 — ASML and Mistral AI today announced a strategic partnership alongside a €1.3 billion investment by ASML. The commitment makes the semiconductor-equipment group the lead investor in Mistral AI's €1.7 billion Series C and links European AI development with some of the world's most demanding engineering systems.

The agreement is designed to apply Mistral AI technology across ASML's product portfolio, research and development, and operations. It does not change the companies' distinct roles: ASML builds lithography systems used by chipmakers, while Mistral AI develops foundation models and software. The partnership creates a route for the two disciplines to work together without turning either company into a subsidiary of the other.

A €1.7 billion round creates a long-term industrial relationship

ASML committed €1.3 billion as the largest participant in Mistral AI's Series C. The round totals €1.7 billion and gives Mistral AI a reported post-money valuation of €11.7 billion. ASML becomes a minority shareholder, giving the partnership strategic weight while Mistral AI remains an independent European technology company.

The figures describe private financing terms, not revenue or guaranteed commercial value. The operational case will depend on whether joint work improves engineering productivity, product performance and the speed with which reliable capabilities reach customers. Both companies will therefore treat deployment quality and measurable use cases as more important than the headline valuation.

A precision lithography machine, wafer inspection stations and an unbranded AI compute rack form a continuous engineering line inside a dark blue cleanroom
The illustration keeps lithography, measurement, engineering validation and computing as distinct stages in one industrial process.

Initial fields for collaboration

  • AI-assisted research and engineering analysis;
  • product capabilities built around controlled technical knowledge;
  • diagnostics, maintenance and operational decision support;
  • software tools that help specialist teams retrieve and assess information;
  • secure deployment patterns suited to sensitive industrial data.

Lithography produces an unusually demanding AI environment

Modern lithography combines optics, mechatronics, materials science, software and nanometre-scale process control. A useful AI system in that setting must work with specialist vocabulary, trace its conclusions to trusted evidence and operate within strict access boundaries. A fluent answer is insufficient if engineers cannot verify its basis or connect it to the physical state of a machine.

Mistral AI contributes models and deployment software; ASML contributes domain knowledge, complex equipment and long engineering cycles. Their joint work can explore where language and multimodal systems shorten searches through technical material, compare observations with known failure patterns or help prepare an analysis for expert review. Final engineering authority remains with qualified teams.

The partnership extends across products, R&D and operations

The announced scope is broader than a single chatbot or internal pilot. Product work may examine how AI supports customers and equipment workflows. Research and development can use models to navigate large bodies of scientific and engineering knowledge. Operational work can focus on planning, service and the structured handling of information.

That breadth requires staged delivery. Each use case needs a defined owner, approved data, evaluation criteria and a rollback path. Systems used for exploration can tolerate different uncertainty from systems that influence maintenance or production decisions. The partners will have to preserve those distinctions as adoption expands.

Industrial data must remain governed

ASML's systems sit in a commercially and strategically sensitive supply chain. Technical records, customer information and equipment telemetry cannot be pooled indiscriminately. Access control, data residency, model hosting, retention and audit trails must be designed into each deployment rather than added after a successful demonstration.

Mistral AI's ability to support controlled deployment is therefore central to the collaboration. The relevant test is not only model quality on a public benchmark, but also whether a system can operate with authorised sources, show useful evidence and respect the boundary between customers, projects and teams.

A European partnership with a practical execution test

The transaction connects two European technology companies at different layers of the computing stack. It provides Mistral AI with capital and an industrial partner, while giving ASML a closer relationship with a model developer. Geographic symbolism, however, does not replace execution.

Progress will be visible in shorter engineering cycles, better-supported decisions, reliable product functions and adoption by specialists who can challenge the output. The €1.3 billion investment establishes the relationship; repeatable improvements in difficult chipmaking work will determine its value.

About Mistral AI

Mistral AI develops foundation models, AI assistants and deployment tools for organisations. The Paris-based company builds products intended for both hosted and controlled enterprise environments.

Company contact
Mistral AI
Company and media inquiries
Website: https://mistral.ai/
Partnership information: https://mistral.ai/customers/asml/