Open WeightMistral

Mistral Large 2

Context

128K tokens

Modalities

text, code

Released

Jul 2024

Overview
Mistral's 123B parameter flagship model positioned as Europe's leading frontier AI. Mistral Large 2 offers strong multilingual performance, long context, and enterprise-focused features with open-weight availability.
Why it matters
Mistral Large 2 is the most capable model built by a European company, and that matters beyond benchmarks. As the EU AI Act reshapes compliance requirements, having a high-quality model from an EU-based vendor simplifies regulatory navigation for European enterprises. The 123B dense architecture offers a different performance profile than MoE models — more predictable latency and simpler deployment. For organizations that need a strong multilingual model with European data governance sensibilities, Mistral Large 2 is the natural choice. Mistral's $2B+ valuation reflects investor confidence in the 'European champion' thesis.

Key strengths

  • 123B dense architecture — strong and predictable
  • Excellent multilingual performance across European languages
  • EU-based vendor simplifies AI Act compliance
  • Open-weight with commercial license
  • Strong function calling and structured output

We cover ai models every week.

Get the 5 AI stories that matter — free, every Friday.

Know the terms. Know the moves.

Get the 5 AI stories that matter every Friday — free.

Free forever. No spam.