Mistral AI

European frontier AI lab (Paris) building open-weight and commercial LLMs, the Le Chat assistant, and an API platform. ~350 staff. ~12B EUR ($14B) valuation, ASML-led 2025 round.

Services Frontier AI / Foundation Models / LLMs
New (0 reviews)
Paris, France (Europe-focused; some remote) HQ
350 employees
2023 founded

What is Mistral AI?

Mistral AI is a European frontier AI lab based in Paris and the most valuable AI company in Europe. It builds large language models, releasing a mix of open-weight models (which anyone can download and run) and commercial ones, including the Mistral Large, Medium, and Small families, the smaller Ministral models, and the Mixtral mixture-of-experts models. On top of the models it offers Le Chat, a consumer and enterprise AI assistant positioned for European users, and La Plateforme, an API and developer platform. Mistral was founded in April 2023 in Paris by three French researchers: Arthur Mensch, the CEO, who came from Google DeepMind, and Guillaume Lample and Timothee Lacroix, who came from Meta's AI research lab where they co-authored the original LLaMA paper. The company positions itself as an open, sovereign European alternative to the large US labs. It employs roughly 350 people. After earlier rounds (a 385M EUR Series A in 2023 and a 600M EUR Series B in 2024 led by General Catalyst), Mistral raised a roughly 2B EUR round in 2025 that valued it at about 12B EUR, or $14B, led by Dutch chip-equipment maker ASML, which took an 11% stake; other backers over time include Andreessen Horowitz, Lightspeed, and Nvidia. Its three founders became France's first AI billionaires.

Mission & values

Put frontier AI in everyone's hands with open, portable, and independent models, building an open and sovereign European alternative in AI.

Qualifications

Mistral AI hires at the frontier of AI: research scientists and research engineers (pre-training, post-training, reinforcement learning, evaluation, multimodal), machine-learning and systems engineers (large-scale training and inference infrastructure, GPUs, distributed systems, performance), plus software engineering (platform, backend, frontend, product), applied AI and solutions engineering, product, design, developer relations, and go-to-market and G&A as it scales commercially. Application flow lives at mistral.ai/careers. The bar for research and ML-systems roles is very high, comparable to the other frontier labs, and deep experience with large-model training, inference at scale, or strong publications helps. The company is Paris-headquartered and Europe-focused, with additional presence in other cities; research and many engineering roles are Paris- or hub-anchored or hybrid, while some roles are remote within Europe. Work authorization in France or the EU matters for many roles, so check each posting.

Leadership

A

Arthur Mensch

Co-Founder & Chief Executive Officer

Co-founded Mistral AI in 2023 and leads it as CEO. Previously a researcher at Google DeepMind. Became one of France's first AI billionaires as the company reached a ~$14B valuation.

G

Guillaume Lample

Co-Founder & Chief Scientist

Co-founded Mistral AI in 2023. Previously at Meta AI research, where he co-authored the original LLaMA paper that helped launch the open-weight LLM era.

T

Timothee Lacroix

Co-Founder & Chief Technology Officer

Co-founded Mistral AI in 2023. Previously at Meta AI research and a co-author of the original LLaMA paper. Leads technology and infrastructure.

Hiring process

  1. 1

    Apply at mistral.ai/careers

    Browse research, ML/systems engineering, software, product, and go-to-market roles and submit through the Mistral careers site.

  2. 2

    Recruiter screen

    Confirms role fit, target level, location and EU work authorization, and expectations.

    About 7 days

  3. 3

    Technical or research interviews

    Research and ML: deep technical interviews on modeling, training, inference, or systems, sometimes with a paper or project discussion. Software: coding and system design. Very high bar.

    About 14 days

  4. 4

    Team and cross-functional interviews

    Conversations with the team and adjacent functions on depth, collaboration, and fit for a fast-moving frontier lab.

    About 10 days

  5. 5

    Offer with equity

    Offer with base plus equity at the current valuation, aligned to role and location.

    About 7 days

Funding

StageSeries C+
Total raised$3,300,000,000
Investors
ASML General Catalyst Andreessen Horowitz Lightspeed Venture Partners Nvidia

Awards & recognition

  • ~2B EUR round at a ~12B EUR ($14B) valuation, led by ASML · 2025

    Mistral AI

  • Founders became France's first AI billionaires · 2025

    Bloomberg

  • Europe's most valuable AI company · 2026

    Mistral AI

Company information

  • Industry: Frontier AI / Foundation Models / LLMs
  • Location: Paris, France (Europe-focused; some remote)
  • Founded: 2023
  • Employees: 350
  • Website: mistral.ai ↗
  • LinkedIn: Company page ↗
  • X / Twitter: Profile ↗

Frequently asked questions

What is Mistral AI?
A Paris-based European frontier AI lab that builds large language models, releasing both open-weight and commercial models (Mistral Large/Medium/Small, Ministral, Mixtral), plus the Le Chat assistant and the La Plateforme API. It is Europe's most valuable AI company.
Who founded Mistral AI?
Arthur Mensch (CEO, ex-Google DeepMind), Guillaume Lample, and Timothee Lacroix (both ex-Meta AI and co-authors of the original LLaMA paper), in April 2023 in Paris. All three became France's first AI billionaires.
How much is Mistral valued at?
About 12B EUR ($14B) after a roughly 2B EUR round in 2025 led by ASML, which took an 11% stake. Earlier backers include General Catalyst, Andreessen Horowitz, Lightspeed, and Nvidia.
Is Mistral remote-friendly?
It is Paris-headquartered and Europe-focused. Research and many engineering roles are Paris- or hub-anchored or hybrid, while some roles are remote within Europe. EU or France work authorization matters for many roles, so check each posting.
What roles does Mistral hire?
Research scientists and engineers (pre-training, post-training, RL, multimodal, evaluation), ML-systems and infrastructure engineers, software engineering, applied AI, product, design, developer relations, and go-to-market. The bar for research and ML-systems roles is very high.
How is Mistral different from OpenAI or Anthropic?
Mistral positions itself as an open, sovereign European lab, releasing open-weight models alongside commercial ones, versus the more closed models of the largest US labs. It is European-headquartered and emphasizes portability and independence.

Stay in the loop.

One email per week, 5 hand-picked roles.