Mistral AI is in the midst of a whirlwind of attention following President Trump’s directives that prompted Anthropic to take its latest AI models offline and the growing call for sovereign technology that reduces dependence on the United States. But the French AI darling is often misunderstood, and the fact that it develops large-scale language models (LLMs) obscures the picture.
Anyone judging Mistral by how close it is to “European OpenAI” will be disappointed. Its chat and agent, Vibe (formerly known as Le Chat), has just an ounce of the brand awareness of ChatGPT, and Claude is more popular than Mistral’s model even among founders based at Station F, a startup campus in Paris.
However, casual observers often miss that the French decacorn company is following Palantir’s strategy, with forward-deployed engineers helping governments and large enterprises deploy AI and tailor it to their use cases.
This approach is also suitable for Mistral means. The company is rumored to raise around $3.5 billion at a valuation of $23.15 billion, almost double its current valuation, but still much lower than Frontier Labs in the US. But its revenue is also increasing. In February, the company revealed that its annual recurring revenue is now over $400 million, up from $20 million just a year ago, and claimed that this year’s ARR is on track to exceed $1 billion.
This has given Mr. Mistral a seat in places like Davos and even chambers like the French parliament, where it is difficult for tech CEOs to get their messages across. Mistral CEO Arthur Mensch has become an ambassador for a particular vision of AI, but there’s still some evangelism to be done when describing his company.
In a lengthy LinkedIn post, Mensch detailed what the Paris-based company does “for a living” – deploying a model and agent platform on the infrastructure of enterprise customers and helping them build custom models using Forge, a platform that allows customers to use their own data for training.
But the misconceptions and high expectations surrounding Mistral didn’t come out of thin air. The company, named after the wind, pursues a grand vision. “We exist to ensure that everyone has access to the best AI systems, outside of the centralized control of nation-states and corporations that feel the need for fine-grained control over AI deployment,” Mensch wrote.
This vision means Mistral is looking beyond the company. The company also aims to continue investing heavily in research to stay ahead of its fundamental AI rivals. Mensch’s post also touches on where he believes the company stands in that regard.
“Currently, we still do not have the best language model, but we have been constantly closing that gap. We have a very exciting model coming this summer, which will be non-discriminatory and will start early access in July. In areas where computing is less constrained, such as speech, vision, and word processing, we have cutting-edge solutions,” Mensch asserted.
Mistral’s next model is already generating buzz at X, with Mensch and Mistral backer Marc Andreessen cracking jokes and fueling memes about the model, which we now know won’t be called “Le Chaton Fat.” This is another sign that the world, especially the “rest of the world,” is paying attention to what Mistral has in his bag.
The most interesting thing may be what happens behind the scenes. Earlier this year, Mistral acquired infrastructure startup Koyeb to further its plans to build a “true AI cloud.” The company also announced a 4 billion euro (approximately $4.56 billion) investment strategy to build data centers in France and Sweden. And the foundations of sovereignty are never far away.
“We are building on the premise that AI technology is a commodity technology that requires secure and affordable supply for all organizations,” Mensch wrote. If you want to know more, keep reading.
Who is the founder of Mistral AI?
Mistral’s three founders share backgrounds in AI research at major US technology companies based in Paris. Before becoming CEO of Mistral, Mr. Mensch worked at Google’s DeepMind. CTO Timothy Lacroix and chief scientific officer Guillaume Lampre are former Meta staff members.
Mistral also awarded co-founding advisor titles to the health insurance startup’s co-founders Alain, Charles Gorintan and Jean-Charles Samulien-Warve (who is also a director). In addition, the company recently appointed three new executives to support its growth: Johann Bergqvist as chief financial officer, Brian Hall as chief marketing officer, and Kamal Brar as senior vice president of partners and alliances.
What are the main models of Mistral AI?
Mistral has developed a wide range of model suites, ranging from LLM to multimodal, inference, audio, and OCR models. Not all models are size-oriented. These include the descriptively named Mistral Small 4 and ‘Les Ministraux’, a family of models optimized for edge devices such as phones. Some are open weight and the code agent Leanstral is now also open source.
What partnerships has Mistral AI entered into?
In 2024, Mistral signed an agreement with Microsoft that included an investment of €15 million and a strategic partnership to distribute the French company’s AI models through Microsoft’s Azure platform.
Mistral announced in May 2025 that it would participate in the creation of an AI campus in the Paris region as part of a joint venture with UAE investment companies MGX, NVIDIA, and French state investment bank Bpifrance.
In June 2025, Mistral announced that it would launch Mistral Compute, an AI-specific platform powered by Nvidia processors, in Europe in 2026. The initiative was hailed as “historic” by French President Emmanuel Macron, who joined Mensch and Nvidia CEO Jensen Huang on stage at the VivaTech conference shortly after the announcement.
In July 2025, Mistral launched AI for Citizens. The initiative, the company said, “can help states and public agencies strategically leverage AI for the benefit of their citizens by transforming public services.”
In September 2025, Mistral and chip company ASML entered into a partnership “to explore the use of AI models across ASML’s product portfolio, as well as research, development, and operations.”
Mistral has also secured strategic partnerships with Accenture, Agence France-Presse, France’s military and employment agency Luxembourg, shipping giant CMA, German defense technology startup Hellsing, IBM, Orange and Stellantis, among others.
How much funding has Mistral AI raised so far?
Most of Mistral AI’s funding to date has been through debt financing, but the company has also raised several rounds of venture funding, totaling about $4 billion, according to Crunchbase.
In June 2023, just one month after its founding, Mistral AI raised a record $113 million in a seed round led by Lightspeed Venture Partners. The largest seed round in European history valued the startup at $260 million, according to people familiar with the matter at the time.
Other investors in this round include Bpifrance, Eric Schmidt, Exor Ventures, First Minute Capital, Headline, JCDecaux Holding, La Famiglia, LocalGlobe, Motier Ventures, Rodolphe Saadé, Sofina, and Xavier Niel.
Six months later, Mistral closed a €385 million ($415 million at the time) Series A, with a reported valuation of $2 billion. The round was led by Andreessen Horowitz (a16z), with participation from Lightspeed as well as BNP Paribas, CMA-CGM, Conviction, Elad Gil, General Catalyst, and Salesforce.
Microsoft’s $16.3 million convertible investment in Mistral as part of the partnership announced in February 2024 was presented as a Series A extension, suggesting the valuation would remain unchanged.
In June 2024, Mistral raised €600 million (approximately $640 million) in equity and debt. The long-rumored round was led by General Catalyst at a $6 billion valuation, with participation from notable investors including Cisco, IBM, Nvidia, and Samsung Venture Investment Corporation.
In September 2025, Mistral completed a €1.7 billion (approximately $2 billion) Series C round led by ASML at a valuation of €11.7 billion (approximately $13.8 billion) with participation from existing backers DST Global, a16z, Bpifrance, General Catalyst, Index Ventures, Lightspeed, and Nvidia.
What companies has Mistral AI acquired?
In addition to infrastructure startup Koyeb, Mistral also acquired Emmi, an Austrian startup focused on physics AI, with ambitions to better support industrial companies’ AI transformation.
Will Mistral AI create its own chips?
Mistral hasn’t designed its own chip yet, but Mensch doesn’t rule out the possibility. “We may end up owning the chip, and I think we should at some point, but for now we’re relying on Nvidia, which is a great partner, and we’re testing a few things here and there,” he told CNBC.
What will the Mistral AI exit look like?
Speaking at the World Economic Forum in Davos in January 2025, Mensch said Mistral was “not for sale.” “Of course, (an IPO) is a plan.”
This makes sense considering the amount of money the startup has raised so far. Even a sale to a rumored potential buyer, like Apple, may not offer enough multiples for investors, not to mention sovereignty concerns for some acquirers.
This story was originally published on February 28, 2025 and will be updated regularly.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
