Parsing

Mistral OCR is a parsing API with agent implications

Mistral OCR turns PDFs and images into ordered text and image output, supports doc-as-prompt workflows, and can return structured data. That makes it more than a prettier OCR endpoint.