Mistral OCR

Mistral AI has introduced Mistral OCR, an API for Optical Character Recognition designed to understand complex documents. The model can extract content from images and PDFs, including text, tables, equations, and embedded media with high precision. It is natively multilingual and benchmark results show it outperforming other leading models like Google Document AI, Gemini, and GPT-4o. The API is available on Mistral's developer platform and can be self-hosted for organizations handling sensitive information.