AI news
Mistral AI Unveils Trillion--Parameter Model "Mistral Large 4"
Date: October 6, 2026
Sources: Mistral AI, Medium
Strategic Overview
Paris--based AI developer Mistral AI launched the public API preview of Mistral Large 4 (officially codenamed "Le Chonk"), marking Europe's first trillion--parameter frontier architecture. Positioned as an open--weight counterweight to proprietary frontier models from OpenAI and Anthropic, the release underscores an intensifying competitive push toward sovereign European AI infrastructure.
Technical & Business Architecture
- Compute & Scaling: The model features 1.05 trillion total parameters configured in a sparse Mixture--of--Experts (MoE) architecture, activating approximately 49 to 52 billion parameters per token. Training was completed natively across 3,800 NVIDIA Grace Blackwell GPUs situated in European data centers.
- Capabilities: Native multimodal processing handles text and vision inputs across more than 160 languages, supporting up to a 1--million--token context window. Benchmark audits indicate specialized enterprise--grade strengths in cybersecurity vulnerability research, automated code generation, and legal reasoning.
- Deployment & Licensing Roadmap: Mistral AI released the preview via its API and scheduled the full public release of the downloadable open weights and commercial licenses for late October 2026, pending final red--teaming with industry partners and security authorities.