MarkPrep vs Pandoc
Pandoc can't convert a PDF to Markdown, so for PDFs and scanned pages use MarkPrep, which converts them in your browser. For Word, HTML, EPUB, LaTeX and dozens of other formats, Pandoc is a free command-line converter that reads them and writes many more.
MarkPrep also converts Word, Excel, PowerPoint and other files, and it estimates how many tokens the result uses in an AI model. You can use MarkPrep to get a PDF into Markdown, then Pandoc to convert it onward.
Side by side
| MarkPrep | Pandoc | |
|---|---|---|
| Reads PDF | Yes, including scanned pages (English OCR) | No, PDF is output only |
| Where it runs | In your browser | Command-line converter on your own computer |
| Install | None, open a web page | Download and install Pandoc |
| Input formats | PDF, Word .docx, Excel, PowerPoint, CSV, HTML, JSON, XML, EPUB, Jupyter, TXT, images, web pages by URL | Word .docx, ODT, RTF, HTML, EPUB, LaTeX, PowerPoint, Excel, Jupyter, CSV and dozens of markup formats |
| Output formats | Markdown, plus Markdown to Word .docx, PDF or HTML | Dozens, including Markdown, Word, HTML, LaTeX, EPUB and PDF |
| Word images | Saved as separate files in a .zip | Saved to a folder with --extract-media |
| Token estimates | For named Claude, GPT and Gemini models | No |
| Chunking for RAG | Chunk preview split at headings, JSONL export | No |
| Licence and price | Free converter; Pro £19 once | GPL version 2 or later, free |
Where Pandoc wins
Pandoc reads and writes far more formats than MarkPrep. It reads RTF, ODT and LaTeX, which MarkPrep doesn't. It writes LaTeX, EPUB and many markup languages, where MarkPrep only writes Markdown, Word, PDF and HTML.
Pandoc is a command-line converter, so it fits into scripts and build steps.
What MarkPrep does differently
MarkPrep reads the text layer of a PDF and runs English OCR on scanned pages, all in your browser. Nothing needs installing.
After conversion, the free AI-Ready Pack gives a token estimate and a context-window check for the model you pick. It also shows a chunk preview split at headings and exports the chunks as JSONL.
Frequently asked questions
Can Pandoc convert PDF to Markdown?+
No. Pandoc can write a PDF, using a LaTeX engine by default, but it has no PDF reader. To turn a PDF into Markdown, use a PDF converter such as MarkPrep, which reads the text layer and runs English OCR on scanned pages.
How do I convert DOCX to Markdown using Pandoc?+
Run pandoc report.docx -t gfm -o report.md. The gfm output format is GitHub-Flavored Markdown. Pandoc reads .docx files only, so save an older .doc file as .docx first.
How do I keep images when converting Word to Markdown with Pandoc?+
Add --extract-media=media to the command. Pandoc then saves the images from the .docx file into a media folder and points the Markdown image links at those files.
How do I stop Pandoc adding line breaks inside paragraphs?+
Add --wrap=none. By default Pandoc wraps output lines at 72 characters, which splits each paragraph over several lines in the Markdown source. With --wrap=none, each paragraph stays on one line.
Can Pandoc convert a whole folder of .docx files?+
If you give Pandoc several input files, it joins them into one document. To get one Markdown file per document, run Pandoc once per file, for example in a shell loop. MarkPrep converts several files at once in your browser.
How do I convert Markdown to HTML with Pandoc?+
Run pandoc -s -o output.html input.md. The -s option makes a complete HTML page with a head and body. Without it, Pandoc writes an HTML fragment. MarkPrep's Markdown to HTML tool does the same in your browser.
Convert a PDF to Markdown in your browser, then take it on to Pandoc if you need another format.