Pixtral Large

Pixtral Large is Mistral's flagship multimodal model, representing the company's most capable offering in the Pixtral family. As a top-tier model from Mistral, it positions the company competitively in the multimodal AI space alongside other major providers. The model supports both text and image inputs with a 131,072 token context window, enabling analysis of lengthy documents alongside visual content. Performance benchmarks show it generates 52.68 output tokens per second with a time to first token of 446 milliseconds, indicating solid inference speed for a flagship multimodal model. Pixtral Large targets applications requiring sophisticated understanding of both text and visual information, competing with other flagship multimodal models in the market. Organizations use it for complex document analysis, visual reasoning tasks, and applications where high-quality multimodal understanding justifies the computational cost of a flagship-tier model.