Mistral AI has released Mistral OCR 4, a new optical character recognition model that can extract and structure content from documents, including text, bounding boxes, block classification, and inline confidence scores. The model supports 170 languages and can run on a single container for self-hosted deployments. It has been benchmarked against leading OCR systems and has achieved high accuracy, especially in low-resource languages. The model is available via API or Document AI and is priced at $4 per 1,000 pages.