Import Word docs and PDFs into Obsidian as clean Markdown

7 August 2026 · Milldown team · All posts

An Obsidian vault is only as useful as what’s actually in it — and for most people, years of notes are trapped in Word files, PDFs, and email attachments. Embedding a PDF in a note doesn’t help: you can’t link to its sections, it doesn’t show up properly in search or graph view, and plugins can’t touch it. The documents need to become real Markdown.

The one-time import

1. Drop your documents folder into Milldown.Word, PDF, PowerPoint, EPUB — mixed folders are fine, and the folder structure is preserved as a mirrored tree of .md files.
2. Run the Clean Up pass.PDF page headers and footers would otherwise litter your notes; the cleanup strips them without touching content.
3. Save the output inside your vault.Obsidian indexes the new files immediately — full-text search, graph view, and backlinks all just work.

Headings become real Markdown headings, so Obsidian’s outline pane and [[Note#Section]] links work on the imported content. Tables arrive as pipe tables you can edit in place. Scanned documents are OCR’d on-device first — see the scanned PDF guide — so even that box of scanned meeting minutes becomes searchable text.

Front matter your plugins can use

Milldown can prepend YAML front matter to every converted file — source filename, conversion date, and a SHA-256 hash of the original. In Obsidian these surface as Properties, which means Dataview queries like “everything imported from the contracts folder this year” work out of the box, and you can always trace a note back to the exact source document it came from.

The ongoing setup: an inbox that converts itself

The nicer trick is making import continuous. Create an Inbox folder anywhere outside your vault and set it as a Milldown watch folder. Every document you drop there — or that your scanner or browser drops there — is converted automatically into its Markdown/ subfolder, which you point into your vault (a symlink works, or just watch a folder inside the vault and let the mirror land there). New PDF in, new note out, no app-switching. More patterns in the batch conversion guide.

A note on links and images

Markdown is a text format: embedded images from Word files and PDFs aren’t carried into the converted note. For image-heavy documents, keep the original file in an Attachments folder and add an embed link in the note — you get searchable text and the visual original, which is usually better than either alone.