No ratings yet
Be the first to review this model
Mistral's frontier-class open-weight multimodal model built on top of Mistral Large 2 with a 1B parameter vision encoder. Excels at understanding documents, charts, and natural images, outperforming GPT-4o on MathVista, ChartQA, and DocVQA. Combines strong text capabilities with advanced visual reasoning for comprehensive multimodal workflows.
Released
November 18, 2024
Parameters
124B
Context
128K
Pricing
Open Source