Mistral AI

Mistral AI is the Paris-headquartered frontier LLM lab founded in **April 2023** by Arthur Mensch (CEO), Guillaume Lample, and Timothée Lacroix — three former DeepMind and Meta AI researchers who built the company on a strategy of open-weight foundational models plus commercial frontier offerings. The product surface spans **Le Chat** (consumer conversational AI hub), **La Plateforme** (developer API), **Studio** (agentic-workflow application platform), **Vibe** (codebase-aware coding models), **Forge** (custom model development), and Applied AI Services for tailored R&D. Customer references include **Stellantis** (automotive), **ASML** (semiconductors), and **CMA CGM** (logistics). Model line-up: Mistral 7B, Mixtral 8x7B / 8x22B, Mistral Small / Medium / Large, Codestral, and specialised vision and embedding models — the open-weight versions (7B, Mixtral 8x7B, Mistral Small) are released under Apache 2.0 and are the de-facto standard for European on-premise LLM deployment. Funding trajectory: seed €105M (June 2023) at €240M led by Lightspeed; Series A US$415M (Dec 2023) at US$2B led by Andreessen Horowitz; Series B US$640M (June 2024) at US$6B; **Series C €1.7B (September 2025) at €11.7B valuation led by Dutch chip giant ASML with a €1.3B cheque for an 11% anchor stake**; March 2026 US$830M debt financing from a consortium of seven banks. Other investors include **NVIDIA** (US), Andreessen Horowitz (US), DST Global (UK-registered with Russian-origin roots), Bpifrance (French state), General Catalyst (US), Index Ventures (UK/US), Lightspeed (US). ASML at 11% makes a European company the largest single shareholder, and Bpifrance provides additional French state alignment — but the US-VC stack collectively still constitutes meaningful US-weighted ownership, which is why the strict-ownership listing sits at `eu_hq_us_funded` rather than `eu_owned`. Procurement-grade picture: the **API product (La Plateforme + Le Chat)** is hosted on cloud-partnership infrastructure that includes Microsoft Azure for European-region deployment — so the at-rest exposure for managed-API customers is material under the strict CLOUD Act stance, and the score caps at 3/5. The **open-weight self-host path** is the cleanest sovereign answer: Mistral 7B + Mixtral 8x7B + Mistral Small running on Hetzner H100, OVHcloud GPU, Scaleway B300, STACKIT GPU, or IONOS sovereign cloud yields an effective 5/5 / none CLOUD Act posture. Best fit: French and EU regulated buyers, defence and public-sector procurement, large enterprises in automotive / semiconductors / logistics, and any organisation evaluating sovereign frontier LLMs — but procurement-led EU-only buyers should choose the self-host path on EU GPU infrastructure rather than the managed API.