[Deprecated] Pixtral Large Mistral AI has launched Pixtral Large, a 124B open-weights multimodal model built on Mistral Large 2, offering frontier-level image understanding and state-of-the-art performance on benchmarks like MathVista, DocVQA, and VQAv2. The model excels at understanding documents, charts, and natural images while retaining Mistral Large 2’s text capabilities, and is available under research and commercial licenses. Concurrent with this announcement, Mistral Large 24.11 receives updates, enhancing long context understanding, system prompts, and function calling for RAG and agentic workflows.

• Pixtral Large, a 124B multimodal model, has been released by Mistral AI.
• It demonstrates frontier-level image understanding, outperforming models like GPT-4o and Gemini-1.5 Pro on benchmarks such as MathVista, DocVQA, and ChartQA.
• The model retains the text performance of Mistral Large 2 and features a 128K context window.
• Pixtral Large is available for research and commercial use under specific licenses.
• Mistral Large 24.11 also receives updates, improving its capabilities for RAG and agentic workflows. Continue reading https://foxvector.com/articles/c0081de5-a951-490c-8005-dd17511d70ae