0xovo/litedoc

LiteDoc is a web page that converts PDF files into Markdown text entirely inside your browser, with no server, account, or installation required. You open the page, drag a PDF file in, and it extracts the text, images, tables, and math from the document locally on your own computer using JavaScript and WebAssembly, then lets you download the result as a zip file containing the Markdown along with the extracted images and tables. The README positions LiteDoc against existing PDF to Markdown tools like Markitdown, Docling, and Marker, which are typically run from the command line or through a backend server and usually need Python installed along with its dependencies. LiteDoc's difference is that it needs no setup at all: since everything runs client side in the browser, files never leave your device, which the README frames as a meaningful privacy benefit compared to uploading documents to someone else's server. Listed features include automatic optical character recognition for scanned pages that have no selectable text, using a WebAssembly build of Tesseract, detection and reconstruction of table structures, extraction of vector based charts as standalone images, support for extracting math formulas as LaTeX and rendering them, right to left language handling for Arabic text, the ability to unlock password protected PDFs locally, and fallback handling for PDFs that use corrupted or custom encoded fonts. To use it, you clone the repository and open its index.html file directly in a modern browser such as Chrome, Edge, Firefox, or Safari, then drag your PDFs into the page. Once conversion finishes, you can preview and lightly edit the generated Markdown in a built in code editor before downloading it. Under the hood it relies on PDF.js for parsing, Tesseract.js for OCR, JSZip for packaging the output, KaTeX for math rendering, and the Ace editor for the in browser preview.