📋About Mistral AI
Updated October 30, 2026Mistral AI is a French AI company founded in 2023 by Arthur Mensch, Guillaume Lample and Timothée Lacroix, former researchers at Google DeepMind and Meta. It is Europe's most valuable AI startup, and its Series D — led by Samsung Electronics with co-leads including EQT-managed Scaleup Europe Fund — is the largest round ever raised by a European technology company, following a Series C led by ASML. Its two lead investors are a semiconductor-equipment maker and a chipmaker rather than software investors, which is the sovereignty pitch expressed as a cap table.
The company develops a full range of foundation models: a large mixture-of-experts flagship, a coding specialist, a mid-sized dense model that self-hosts on as few as four GPUs, compact models under Apache 2.0, and an edge-deployment family. Alongside them sit Voxtral, a speech line whose October 2026 release adds transcription, expressive speech and a kit for building voice agents (the original speech model's downloadable weights are licensed for non-commercial use only), and a document-intelligence model that extracts structured text, tables and layout from scans across 170 languages in a single self-hosted container, positioned as the ingestion layer for enterprise search and retrieval. Timestral, its first time-series family, forecasts sensor, financial and operations data on ordinary processors; its weights are free only for companies under 20 million dollars in monthly revenue.
Vibe is the unified consumer and enterprise platform, combining chat, productivity and coding agents under one license — planning multi-step tasks across Google Workspace, Outlook, SharePoint, Slack and GitHub, and running coding agents in isolated cloud sandboxes launched from the web, a CLI, a VS Code extension or third-party apps. It is priced below a single competing subscription from the major US labs.
Mistral positions itself as Europe's answer to the US labs, pairing frontier capability with open-source commitment and EU data sovereignty. It runs its own compute — a Paris data center, an inference facility south of Paris, a planned Swedish site for sovereign EU inference, and a large Grace Blackwell training cluster — with an acquisition adding AI cloud infrastructure. Defense contracts with Helsing and the French military demonstrate government trust in European AI sovereignty.
Two expansions widen that story. Through the acquisition of the Austrian startup Emmi AI, Mistral launched Physics AI, foundation models that learn from physics-solver outputs and predict the behavior of physical systems in seconds on a single GPU, replacing simulations that traditionally take hours to weeks — with Airbus, ASML, BMW Group, Safran and Siemens Energy as disclosed customers, reframing the roadmap beyond chat and coding into digital twins for high-stakes manufacturing. And a collaboration with HUMAIN, the Saudi AI company majority-owned by the Public Investment Fund, extends the sovereignty pitch beyond Europe for the first time, spanning regional infrastructure and Arabic-strong model development. The significance is that an argument built for European buyers travels to any state willing to fund its own stack, which widens the addressable market while tying Mistral to a sovereign backer.
🛠️Products & Tools (11)
Open-source terminal coding agent powered by Devstral 2 with custom subagents (TOML config), slash-command skills, MCP support, and on-premise deployment.
Mistral's 3 billion parameter open-weights safety classifier. Judges text and images against moderation policies written in plain language at inference time, under Apache 2.0.
Mistral's flagship open-weight model: 1 trillion parameters (49 billion active), 1 million-token context, multimodal. Weights released October 29, 2026 under the Mistral Ecosystem Open-Weight License, free for most uses but requiring a commercial license for companies that resell API access. Pitched for agentic coding and defensive security.
Mistral's coding specialist. 123B params scoring 72.2% on SWE-bench Verified, claims 7x more cost-efficient than Claude Sonnet. Devstral Small 2 (24B, Apache 2.0) scores 68%. Includes Vibe CLI.
Efficient MoE model — 119 billion total parameters, 128 experts, approximately 6.5 billion active per token, 256,000 token context, Apache 2.0 license
Structure-aware document-intelligence OCR model — extracts text, tables, and layout across 170 languages, self-hostable in a single container, at $4 per 1,000 pages. Leads the OlmOCRBench benchmark.
Mistral's open Apache-2.0 model for formal mathematical proof in Lean 4 — a mixture-of-experts specialist that aces the miniF2F benchmark and finds real software bugs.
Data-driven foundation models for industrial engineering — learning from physics-solver outputs to predict the behavior of physical systems in seconds on a single GPU. Launch partners include Airbus, ASML, Safran, and Siemens Energy.
Mistral AI's consumer AI platform. Powered by Mistral Large 3 (675B MoE). Mobile apps, Deep Research mode, Projects, code interpreter, Flux Ultra image gen. EU data sovereignty. Pro at €14.99/month.
Mistral AI's first embodied-AI model — an 8 billion parameter vision-language-action model that steers robots through unfamiliar environments using only a single RGB camera, with no LiDAR, depth sensors, or pre-built maps. Reports 76.6% success on unseen Room-to-Room continuous-environment navigation benchmarks; hardware-agnostic across wheeled, legged, and flying robots. Announced July 2026.
Mistral AI's text-to-speech line: a 4 billion parameter model from March 2026 whose downloadable weights are licensed for non-commercial use only, and Voxtral 3 (October 2026), with speech in 35+ languages, transcription and a voice agent kit through Mistral's API at $13 per million characters.
📦Earlier generations & discontinued (1)
Kept as a reference. Each page links to the version that replaced it.
Mistral's previous flagship open-weight model: 675 billion-parameter MoE (41 billion active), 256K context, multimodal, modified MIT license. Succeeded in October 2026 by the 1 trillion-parameter Mistral Large 4, which ships under a different license; Large 3 remains the smaller option to self-host.
Keep track of the companies you’re watching
- Save the companies you want to follow
- Get ⚡ alerts when your saved companies change
- Every product they ship, cross-linked to 1,000+ AI tool profiles
- Today’s top AI Stories — the day’s most important AI news, free
Swipe for Recommended for you and My AI Tools
Your AI Hub — sample data. See desktop view example



