This project is based on a visual model API service that provides the functionality to convert PDF documents into Markdown. The principle involves first converting the original file (such as pdf, docx) into an image, then sending the image to the visual model for processing, and finally summarizing all the results to generate a complete Markdown file.