Samsung weighs up to €1 billion Mistral investment

Samsung is discussing an investment of as much as €1 billion in Mistral, the French AI model developer, in a deal that would value the company at roughly €20 billion, the Financial Times reported. If completed, the investment would give one of the world’s largest memory chip makers a deeper financial position in a European AI company that is also expanding its cloud and compute ties with Microsoft.

The talks come less than a year after Mistral was valued at €12 billion. Samsung had already backed the company in 2024 through its venture arm, according to earlier reporting, and Swedish investor EQT is also expected to take part in the new financing, the Financial Times reported.

The reported round has not been announced by Mistral or Samsung, and the final size, ownership stake, governance terms and closing timetable were not disclosed. The headline valuation would still be notable for Europe’s AI market, where few model companies have attracted capital at levels comparable to the better-funded U.S. labs.

Samsung’s AI exposure is spreading beyond memory

Samsung’s interest in Mistral sits alongside its broader push into AI infrastructure. The company is a major supplier of memory chips, a key input for AI servers, and has also been in talks with Anthropic about manufacturing a custom AI chip, according to prior reporting.

That makes the Mistral talks more than a standard venture bet. Samsung sells into the AI buildout and is trying to position itself closer to the customers and model developers driving demand. A financial stake in Mistral would give Samsung another link to a company buying and deploying large amounts of compute, though there is no reported manufacturing or supply agreement tied to this potential investment.

Mistral is raising against ambitious revenue claims

Mistral CEO Arthur Mensch said in February that he expected the company’s annual revenue to exceed $1 billion by the end of the year. The company did not disclose a current revenue figure in connection with the reported Samsung talks, so that projection remains difficult to benchmark from the available information.

Mistral has also broadened its relationship with Microsoft. This week, the companies expanded their partnership, with Microsoft committing to spend billions of dollars on Mistral’s compute infrastructure. Mistral is increasing its GPU capacity with thousands of Nvidia Vera Rubin chips and is making its models available through Microsoft Azure.

The Azure distribution includes availability in air-gapped environments, meaning deployments without internet connectivity, for regulated sectors such as finance and healthcare. That is a practical route to enterprise workloads that cannot use ordinary public-cloud AI services, although the companies have not disclosed how much revenue those deployments are expected to generate.

For Mistral, the reported Samsung investment would add another large industrial name to its cap table as it tries to scale model development, distribution and infrastructure without relying solely on U.S. cloud providers. For Samsung, it would extend an AI strategy that already spans memory, potential custom chip manufacturing and venture investments in model companies.

This story draws on original reporting from The Decoder.