Can Obsidian Convert PDF to Markdown?
Obsidian can't convert a PDF to Markdown on its own. It opens PDFs in a built-in viewer and lets you embed or link them, but the file stays a PDF. The nearest thing to PDF to Markdown in Obsidian is copying text out of that viewer by hand. To get a whole document into a Markdown note, convert the PDF outside Obsidian and save the .md file into your vault folder.
This works because a vault is an ordinary folder of Markdown files on your computer. Obsidian picks up any file you add to that folder.
What Obsidian does with a PDF
Obsidian lists PDF among the file types it supports, next to Markdown, images, audio and video. For a PDF, support means Obsidian can open the file and show it.
With a PDF in your vault, you can:
- open it in a tab, with zoom, thumbnails and the PDF's own table of contents;
- embed it in a note with
![[Report.pdf]], or jump to one page with#page=Nafter the file name; - copy a link to a page, or to a passage you have selected;
- search inside it on desktop with
Ctrl+Fon Windows and Linux orCommand+Fon macOS; - select text on desktop and copy it, or copy it as a quote with a link back to the passage.
Copying works for a paragraph or two. For a whole report it is slow, and you still have to add the headings and tables by hand. Obsidian's help also says the app can't annotate or highlight a PDF.
Obsidian's official Importer plugin brings in notes from apps such as Notion, Evernote, Apple Notes and OneNote. It also imports HTML, CSV and Markdown files. PDF and Word files are not on its list. Community plugins can add other file formats, but they are written by third parties, so check each one before you trust it with private files.
Not every PDF needs converting. A form, a slide deck or a scanned book you only read can stay a PDF, embedded in a note where you write about it. Convert the PDFs whose text you want to search, quote, link to by heading or hand to an AI model.
Is there a Markdown converter for Obsidian?
MarkPrep is a Markdown converter that works alongside Obsidian. It runs in your browser, so the PDF is not uploaded, and it writes a .md file for your vault. To move a PDF to Obsidian as a note:
- Go to MarkPrep's PDF converter and drop your PDF onto the page.
- Turn on "YAML front-matter" in the options. Each file then starts with front matter: its title, source file name, page count, word count, token estimate and date.
- Leave the clean-up options on. They remove headers and footers that repeat on most pages, page numbers on their own line, and words hyphenated across a line break.
- Read the Markdown tab and fix anything odd, such as a heading at the wrong level.
- Download the
.mdfile and move it into your vault folder, or into a subfolder of it.
Obsidian refreshes the vault when files change outside the app, so the note appears in the file explorer without a restart. The front matter shows up as the note's properties, because Obsidian stores properties as YAML at the top of the file.
Keep the original PDF in the vault next to the note. Then add ![[Report.pdf]] at the end of the note, and the source sits one scroll away when you need to check a figure.
What a converted PDF looks like in your vault
Here is the top of a note made from a PDF called Annual report 2026.pdf, with front matter turned on. The field names are the ones MarkPrep writes. The title comes from the file name, the date is the day you convert, and the numbers are an example.
---
title: Annual report 2026
source: Annual report 2026.pdf
pages: 48
words: 18250
tokens: 41200
date: 2026-11-01
---
# Annual report 2026
## Chair's statement
Larger text in the PDF becomes Markdown headings, so Obsidian's Outline view shows the report's sections. Tables are rebuilt as Markdown tables when their columns line up on the page. Check them against the PDF, because a wrapped label or a table split across two pages can break a row.
Some PDFs need more care. A page laid out in two columns comes out with lines from both columns mixed, because the converter reads straight across. For a scan, the local OCR handles English text only and is less accurate than a PDF with real text. The Verify tab says when OCR was used and lists pages it read with low confidence. Mixed columns show up only when you read the text, so read the note once before you link to it.
Word and DOCX to Markdown for Obsidian
Word files take the same route. Drop a .docx file onto the Word converter page. Word headings become Markdown headings, and lists and tables are kept.
Images need one extra step. When a Word document has pictures, MarkPrep saves them as separate files and offers a .zip download. It holds the .md file and an images folder that the note links to. Unzip it into your vault with the images folder beside the note, so each link still points at its picture.
Older .doc files have to be saved as .docx in Word first. A Google Doc can be downloaded as a .docx file and converted the same way.
Other files for your vault, and notes going out
MarkPrep has converters for other source files you may want in a vault. Web pages go through URL to Markdown, and saved HTML files through HTML to Markdown. Spreadsheets become one Markdown table per sheet with Excel to Markdown. To publish a note on a website, Markdown to HTML turns it into a web page file. Wikilinks such as [[Note name]] and embeds such as ![[Report.pdf]] are Obsidian's own syntax, so they stay as plain text in the HTML. Change them to standard Markdown links first. To share a note as a PDF instead, use Obsidian's own export. On desktop it saves a note as a PDF, with a choice of page size and margins.
For a whole folder of PDFs, MarkPrep Pro adds a command-line converter, sold once for £19. Its watch command keeps an eye on one folder and converts each file you drop into it to Markdown, saved in a second folder. Set that second folder inside your vault, and each converted file appears as a note. It handles PDF, Word, Excel, HTML, CSV, JSON and text files. It has no OCR, so scanned PDFs still go through the browser.
Start with one file. Convert it on the PDF page, move it into your vault, and compare it with the PDF. The Obsidian page has front matter turned on from the start.
Try it on your own file
Convert a document and watch the token counter. It's free, needs no account and uploads nothing.