Mistral raises €3 billion to build AI that companies can keep behind their own walls

Key points

• Mistral AI raised €3 billion in a Series D funding round, valuing the company at more than €21 billion.
• Samsung Electronics led the round, the largest equity fundraising ever completed by a European technology company.
• Mistral was founded three years ago and now serves more than 125 large enterprises across 20 countries, including Airbus, ASML and HSBC.
• The company builds open-weight models, meaning the underlying code is publicly available, so businesses can run the AI on their own servers rather than a third party's.
• New investors include Advent, BlackRock-managed funds and the Grand Duchy of Luxembourg.

Three years old and already carrying the biggest fundraising cheque in European tech history. That's where Mistral AI, the Paris-based artificial intelligence company, stands after closing a €3 billion Series D round at a valuation above €21 billion.

Samsung Electronics led the investment, joined by co-leads Scaleup Europe Fund, managed by private equity firm EQT, and existing backer PSG Equity.

What does Mistral actually do?

Mistral builds large language models, the same type of technology that powers chatbots like ChatGPT, and sells them to large businesses and governments. Its focus is open-weight models: AI systems where the underlying code and trained parameters are published openly, so organisations can download them and run them on their own computers, with no data leaving their building.

That matters enormously for organisations handling sensitive information. A closed model forces you to send your data to someone else's server every time you ask a question. An open-weight model doesn't.

Mistral describes this as "sovereign AI", meaning AI an organisation controls completely: data stays inside its own walls, the model can be customised, the computing hardware is private, and every decision the system makes can be audited.

Why is the funding so large?

The money will go toward frontier research (developing the most powerful AI systems), computing infrastructure for training those models, and commercial expansion. Mistral already operates across 20 countries and supports more than 125 large enterprises on what it calls mission-critical AI, meaning AI woven into core business processes rather than used as a novelty. Clients named in the announcement include aircraft-maker Airbus, chip-equipment company ASML and bank HSBC.

With its Series C led by ASML and now a Series D led by Samsung, Mistral has pulled in two of the world's most demanding industrial manufacturers in successive rounds. We've been covering Mistral AI since 4 August 2026, and this pattern of industrial backers is the clearest sign yet that complex, regulated industries are choosing data control over raw model power.

Existing backers include venture capital firms Andreessen Horowitz, Index Ventures and General Catalyst, plus corporate investors Nvidia and Salesforce Ventures. New arrivals are Advent, BlackRock-managed funds and the government of Luxembourg.

What does this mean for ordinary people?

If you work in healthcare or finance, AI tools built on Mistral's models may reach you through your employer's systems rather than a consumer app. The selling point your employer is paying for: your patient records and financial data aren't used to train someone else's model.

For the broader AI market, this round signals that European and Asian investors are placing serious money behind an alternative to dominant US providers. Whether that produces genuinely different products for end users will depend on what Mistral ships next.

Common questions

Is open-weight AI less powerful than the models from OpenAI or Google?

Not necessarily. Open-weight models have closed the gap in recent years. The trade-off is between raw performance on the hardest tasks versus control and privacy, and Mistral argues its models are strong enough for most enterprise needs.

Does this funding mean Mistral products will change for current users?

Mistral says the money will expand research and infrastructure rather than shift its business model. Its open-weight models are expected to remain available, and the company says customers won't be locked into a single pricing structure or roadmap.