Mistral Large 3 (675B Instruct 2512)

Mistral Large 3 is a multimodal Mixture-of-Experts model with 41B active parameters and 675B total parameters, trained from scratch on 3,000 H200 GPUs. This is an FP8 instruct version fine-tuned for instruction tasks, ideal for chat, agentic, and instruction-based scenarios. Designed for reliability and long-context understanding, suitable for production-grade assistants, RAG systems, scientific workloads, and complex enterprise workflows.