Mistral AI closes €600M Series B

Mistral AI, the Paris-based large-language-model developer founded by former Meta and Google DeepMind researchers, closed a €600 million (roughly $640 million) Series B round led by General Catalyst. TechCrunch reported the round was split between €468 million in equity and €132 million in debt, and valued the company at around $6 billion — a large jump from the roughly $112 million seed round the company had raised about a year earlier. Participants included returning investors Lightspeed Venture Partners and Andreessen Horowitz alongside a broad group of new corporate and strategic backers, among them Nvidia, Samsung, Salesforce Ventures, BNP Paribas, Cisco, IBM and ServiceNow.

Chief executive Arthur Mensch said the funding “guarantees the company’s continued independence,” which the company and its investors framed as remaining under founder control rather than being absorbed by a larger US platform company, even as several of the round’s participants were themselves major US technology firms. The round came as Mistral positioned itself against OpenAI’s GPT-4o, Anthropic’s Claude 3 and Meta’s Llama 3, offering both openly licensed models under Apache 2.0 and proprietary models, including Mistral Large, available through an API, alongside a consumer chat product, Le Chat, and a distribution partnership with Microsoft Azure.

The raise made Mistral the best-funded European AI lab by a wide margin and reinforced its position as the leading non-US, non-Chinese entrant among frontier-adjacent model developers — a status it used over the following two years to argue for European AI sovereignty in policy debates over the EU AI Act and public-sector procurement.