Mistral 3

Mistral 3 offers a series of advanced models for various computational needs, featuring both dense and sparse mixture-of-experts architectures. Key characteristics include: • Three distinct small, dense models (3B, 8B, 14B parameters) • A highly capable sparse mixture-of-experts model (Mistral Large 3) with 675B total parameters • Released under the Apache 2.0 license for broad accessibility • Strong performance-to-cost ratio across its model family • Advanced instruction-tuned capabilities for complex tasks This suite of models provides a versatile foundation for a wide array of applications. The smaller 'Ministral' models are engineered for efficiency, offering excellent performance relative to their computational demands, making them ideal for scenarios where resource optimization is critical. Mistral Large 3, in contrast, is designed for cutting-edge performance, capable of handling highly sophisticated instructional prompts and delivering high-quality outputs. Built with a focus on open access, these models empower developers and organizations to integrate advanced textual and multimodal understanding into their platforms. They support diverse applications, from enhancing user interaction to sophisticated content generation and data interpretation. The Apache 2.0 license encourages broad adoption and customization. These models are perfect for developers, researchers, and enterprises seeking robust, open-source building blocks for applications that require advanced language understanding, generation, and multimodal processing. They cater to a spectrum of uses, from resource-constrained edge deployments to high-performance cloud environments.