French AI startup Mistral AI is reportedly preparing to raise an additional $1 billion, aiming for a valuation of $10 billion. Founded by former Meta and Google DeepMind engineers, the company focuses on open-weight, modular AI models and has attracted global investors like Nvidia and Andreessen Horowitz.
Launched by Arthur Mensch, Guillaume Lample, and Timothée Lacroix, all École Polytechnique alumni, Mistral aims to make AI accessible and aligned with European data norms. Mensch contributed to DeepMind’s “Chinchilla” model, while Lample was pivotal in Meta’s LLaMA models. The startup quickly gained traction with a €105 million Seed round and a €600 million Series B raise, leading to a €5.8 billion valuation as of June 2024.
Mistral’s products, Mistral Large and the Mixtral series, are strong contenders in the AI market, challenging U.S. giants like OpenAI. Its open-source, GDPR-aligned approach has attracted clients like BNP Paribas and AXA. Le Chat, its chatbot, surpassed one million downloads in two weeks, aided by Free Mobile bundling.
Mistral has formed key partnerships, with Microsoft investing €15 million and distributing its models via Azure. CMA CGM committed €100 million to integrate its technology. The company plans to use new capital to build a data center near Paris, including an €8.5 billion facility in Essonne powered by nuclear energy.
Mistral’s models appeal to sectors needing transparency and compliance, such as healthcare and finance. With nearly 400 employees expected by mid-2025, the leadership reflects inclusion, with women holding 50% of senior roles and a diverse workforce. Mistral was recently featured in EU-Startups’ article highlighting Europe’s shift toward vertical AI specialisation.