Mistral started operations a year ago, promising to compete with OpenAI from Paris. Today, it has a flurry of good news to share, capped by the investment round it just closed, which values it at roughly $6.5 billion.
Why it matters. Mistral is Europe’s big chance to have its own leader in generative AI, a company capable of competing with the big players from the U.S. and China. Its open-source approach and multilingual domain separate it from competitors like OpenAI, which has the support of Microsoft.
For France and the European Union, supporting Mistral is critical to reducing dependence on large U.S. platforms and having a role in defining the standards for the technology that will transform society.
Mistral’s trajectory. Arthur Mensch, Timothée Lacroix, and Guillaume Lample, former engineers from Google's DeepMind and Meta, founded the company in June 2023. In their first year, they launched AI models in French, English, German, Italian, and Spanish that rival GPT-4.
The three engineers have raised more than $1.1 billion in three rounds of funding:
- The one from a year ago, for about $115 million (a European record for a seed round).
- The one in December 2023, for nearly $420 million, brought the company’s value to $2.17 billion.
- The June 2024 round was for $650 million, which values Mistral at almost $6.5 billion.
Microsoft also invested $16 million in February to incorporate Mistral models into Azure but didn’t participate in the latest round. Investors include Nvidia, Samsung, IBM, BNP Paribas, Salesforce, Andreessen Horowitz, and Lightspeed, among others.
Mistral’s founders. Image | Lightspeed
Support from the French government. President Emmanuel Macron called Mistral an example of “French genius” and invited its CEO to dinner at the Elysée Palace. Finance minister Bruno Le Maire frequently praises the startup as a national tech gem.
Former digital affairs minister Cédric O is an advisor and shareholder in Mistral. Notably, France also lobbied in Brussels on the new EU AI law, passed a few months ago, to limit the regulation of open-source AI systems, which benefits Mistral.
Challenges and opportunities. Mistral is committed to releasing the code of its models so that anyone can examine, use, and adapt them, something it believes will accelerate innovation.
However, keeping up with the pace of computational investment required by these systems is a significant challenge. Mistral says that “just over 1,000 high-end GPUs” are enough to train its models.
What’s next? Mistral's next goal is to make its technology profitable and scalable. It makes several million a year with 60 employees, boasting clients like Renault and BNP Paribas, and its goal is to help companies use AI to reinvent themselves.
Another asset is its capability to process language and interfaces in different European languages, unlike its rivals, which are limited to English.
Outlook. Mistral’s rise in just one year demonstrates that Europe can fight in the age of generative AI but also shows its dependence on venture capital and U.S. cloud providers.
If Mistral succeeds, it will be the new technology standard in Europe. If it fails, it will be a missed opportunity for a continent whose reputation for over-regulation and lack of innovation grows each year.
Images | Wikimedia Commons | Mistral | Xataka On
Related | Nvidia Has Just Surpassed Apple as the World’s Second Most Valuable Company
View 0 comments