Mistral AI

Mistral AI is a Paris-based frontier AI research lab developing open-weights and enterprise language models. Founded by former AI researchers from Meta and Google DeepMind, Mistral pioneered breakthrough sparse Mixture-of-Experts (MoE) architectures like Mixtral 8x7B and Mistral Large. Mistral provides high-efficiency, customizable AI models tailored for privacy-conscious enterprise deployments. Backed by General Catalyst, Lightspeed, Andreessen Horowitz, and Nvidia.