Mistral’s €3 billion open AI bet: Why Samsung, ASML are backing it and what’s next for French startup?

French artificial intelligence company Mistral AI has raised €3 billion in fresh funding at a post-money valuation of more than €21 billion, giving the three-year-old startup substantially more capital to develop advanced AI models, expand its computing infrastructure and strengthen its position as Europe’s homegrown alternative to US technology giants.

The Series D round, which Mistral described as the largest equity fundraising by a European technology company, was led by Samsung Electronics, the EQT-managed Scaleup Europe Fund and existing investor PSG Equity. New investors include Advent, funds and accounts managed by BlackRock and the Grand Duchy of Luxembourg, alongside a long list of existing backers.

Why Mistral’s €3 billion funding matters for Europe

Mistral’s latest fundraising comes as governments and businesses increasingly seek greater control over where their AI systems, data and computing infrastructure are hosted.

The Paris-based company has positioned sovereign AI as a central part of its strategy. Its approach combines open-weight models with infrastructure and products that allow organisations to deploy and customise AI within their own environments.

For European governments in particular, the strategy addresses concerns around dependence on foreign technology providers and the potential disruption of access to overseas AI systems.

"The fact that the EU or Europe have to have their own kind of AI provider in the game is important," Mistral CFO Johan Bergqvist said.

Mistral's models can be downloaded and customised by customers, giving enterprises greater control over their data and intellectual property. The company says more than 125 global enterprises now use its technology, including Airbus, ASML and HSBC.

Mistral plans bigger AI models and more computing capacity

The new capital will also support Mistral’s push to expand its computing footprint. CEO Arthur Mensch said the company plans to increase the amount of compute capacity it owns, including through its own data centres, while continuing to rent computing resources.

"Long term, the plan is to fully rely on capacity that we are building ourselves, and so that means that the amount of compute that we own is going to grow ... around 100% in the next five years," Mensch said.

He also said the company plans to train "bigger and faster models."

Mistral has a data-centre project planned in Sweden and is seeking to expand its infrastructure as demand for its models grows. The company expects annual recurring revenue to exceed $1 billion this year.

Open-weight AI is Mistral’s answer to closed models

While OpenAI and Anthropic have built businesses around proprietary AI systems, Mistral has made openness a core part of its commercial strategy.

Open-weight models allow customers to access and modify the underlying model, potentially making it easier for companies and governments to deploy AI on their own infrastructure.

"We see that open source is resonating more and more with corporates that want to be in control of their own sovereignty, their own AI transformation and applications, and they also want to be more careful and cautious about with whom and where they share their confidential customer data as well as their own IP data," Bergqvist said.

Mistral argues that this model can reduce dependence on a single provider’s pricing, product roadmap or availability.

Mistral faces OpenAI, Anthropic and Chinese AI competition

Despite its rapid growth, Mistral remains far smaller than US rivals such as OpenAI and Anthropic. It also faces increasing competition from Chinese AI developers, whose open models have become more capable.

Mensch said Mistral’s forthcoming models would be "very competitive", while highlighting the importance of continuing to develop its own technology to provide customers with a predictable path to newer models.

"Now we also need to be a trusted partner for our customers, and they want us to be sure to certify that in one year from now they will get access to better models than they have access to today, and the only way we can provide that guarantee is by continuing to train our models ourselves," Mensch said.

For now, Mistral’s strategy is to combine advanced model development with control over infrastructure, giving Europe a company capable of competing in AI while keeping a stronger degree of technological independence.