Mistral AI Raises €3B Series D Led by Samsung

Mistral AI closed a 3 billion euro Series D funding round on September 8, 2026, the largest equity raise ever completed by a European technology company, according to the company's own announcement. Samsung Electronics led the round, with the EU-backed Scaleup Europe Fund, managed by EQT, and existing investor PSG Equity co-leading. The deal pushes Mistral's post-money valuation past 21 billion euros, nearly double the 11.7 billion euros it was worth after its Series C roughly a year earlier. For a company most people still describe as “Europe's answer to OpenAI,” the list of names behind this check tells a different story.

Who actually wrote the checks

Samsung's exact contribution has not been broken out publicly, but it is described as the lead investor alongside the two co-leads. Scaleup Europe Fund's participation is notable on its own: it is a roughly 5 billion euro, EU-backed vehicle managed by EQT, and this is reportedly its first investment since launching, aimed squarely at keeping strategic technology capacity inside Europe. New money also came from Advent, funds and accounts managed by BlackRock, and the Grand Duchy of Luxembourg. Returning backers include a16z, ASML, General Catalyst, Lightspeed, NVIDIA, and Salesforce Ventures, all of whom had already bet on Mistral in earlier rounds. That mix, a Korean electronics giant, a Brussels-backed fund, a US private equity firm, and a roster of existing venture investors, is unusual even by the standards of a sector used to unusual cap tables.

Why this round breaks the usual script

Most AI mega-rounds this size have followed a familiar pattern: a US hyperscaler puts in cash and cloud credits tied to a compute-purchase commitment, effectively buying training capacity back from itself. Microsoft's stake in OpenAI and Google's position in Anthropic both fit that mold. Mistral's round does not. Its lead investor is a memory and semiconductor manufacturer with no cloud business to sell Mistral back into, and its second-largest new backer is a public EU fund built explicitly around sovereign infrastructure, not commercial cloud economics. That combination is the real story here for anyone tracking where European AI money actually flows: capital arriving from hardware and sovereign-fund channels rather than hyperscaler balance sheets.

What the 3 billion euros is actually funding

Mistral says the priority is compute it owns outright rather than compute it rents. CEO Arthur Mensch told reporters the plan is to build and operate its own data centers, expanding European capacity to 200 megawatts by the end of 2027, while continuing to rent additional capacity in the meantime. “Long term, the plan is to fully rely on capacity that we are building ourselves,” Mensch said, adding that compute the company owns is set to grow by roughly 100 percent over the next five years. Mensch also said Mistral expects to cross 1 billion dollars in annual recurring revenue before the end of the year, a figure that, if it holds, would mark a sharp jump in the company's commercial traction alongside the funding news.

Why builders should care

For founders and teams choosing model providers, this round is a signal, not just a headline number. A well-capitalized Mistral with owned European data centers changes the calculus for anyone weighing data residency, EU regulatory comfort, or plain diversification away from a small handful of US-hyperscaler-backed labs. It also shows that serious AI infrastructure money is not exclusively a Silicon Valley story anymore: sovereign-aligned funds and Asian hardware manufacturers are now writing checks at a scale that used to be reserved for Microsoft, Google, and Amazon. Compare the scale here to a round like Lovable's recent Series C, and the gap illustrates just how much capital concentration is happening at the frontier-model layer specifically, even as smaller AI-native companies raise real money of their own. Anyone building a roadmap around model choice, and anyone simply trying to keep up with AI news week to week, now has one more reason to watch what Mistral ships next, and where it puts those new data centers. None of the figures above change what Mistral's models cost to call or how they benchmark today. What they change is the runway: a company with billions in owned compute behind it can afford to keep shipping frontier-class models on a European footprint for years, rather than one or two more funding cycles.