markpdfdown

Why we included this project

MarkPDFDown cuts out the tedious copy-and-paste work of moving documents from PDF to an editable format. It feeds each page to a vision-capable language model and returns clean Markdown, so tables, formulas, and diagrams come through better than they do with rules-based extraction libraries. Headings and lists stay intact, the CLI accepts both files and piped input, and you can restrict conversion to a specific page range, which makes it a good fit for batch jobs. It needs an OpenAI or OpenRouter API key, and there's also a desktop app for anyone who prefers a GUI over the command line.