PDF to Markdown Converter
Turn a PDF into a .md file with headings, lists and tables written as Markdown syntax, ready for a notes vault, a docs repo or a retrieval pipeline.
PDF Converter
or drop PDF files here ·
Max 95 MB per file· Up to 20 files· Deleted after 30 minutes
Files are processed on our servers and deleted after 30 minutes.
Scanned PDFs are pictures of pages. Auto runs OCR only on pages without a text layer; On forces it for every page.
The language printed in the scan. Picking the right one noticeably improves accuracy for accents and non-Latin scripts.
Flowing text reflows into normal paragraphs that are easy to edit. Exact layout places every block where it was, which looks identical but is harder to edit.
Where detected tables go in the workbook.
Page numbers or ranges separated by commas, e.g. 1-3, 5, 8-10.
How to use this tool
- 1
Add the PDF
Drop up to 20 PDFs of up to 95 MB each. Scanned pages are read with OCR while Text recognition (OCR) stays on Auto.
- 2
Choose images and pages
Keep images is on, so figures come along. Turn it off for a text-only file, and use Pages to convert to skip a cover or appendix.
- 3
Convert and check the outline
Press Convert to Markdown and download the .md file, or a ZIP for several. Glance at the heading levels first.
A PDF stores glyphs at positions; Markdown stores structure as characters: # for a heading, a dash for a list item, pipes for a table. Converting reads that structure back from font sizes, indents and columns.
How PDF to MD conversion rebuilds headings, lists and tables
Most PDFs do not know that a line is a heading. Unless the file was exported as a tagged PDF, a feature added in PDF 1.4 in 2001 and required by the PDF/UA accessibility standard (ISO 14289), a title is just text in a larger or heavier font. The converter ranks the type styles it finds and maps the largest recurring ones to #, ## and ###, so a report with a clear type hierarchy comes out with a clean outline.
Lists are recognized from markers and indents: bullets become -, numbered lines 1., deeper indents nested items. Text aligned in columns becomes a pipe table, a syntax GitHub formalized in its GFM specification in 2017; the Markdown John Gruber published in 2004 had no tables at all. Links keep their targets as [text](url).
Line breaks that only existed because a line hit the margin are joined back into paragraphs. A running header or page number landing between two paragraphs is the usual thing to delete afterward.
| from ↓ · to → | MD | TXT | DOCX | HTML |
|---|---|---|---|---|
| Headings | Headings to MD: yes | – Headings to TXT: no | Headings to DOCX: yes | Headings to HTML: yes |
| Bulleted and numbered lists | Bulleted and numbered lists to MD: yes | – Bulleted and numbered lists to TXT: no | Bulleted and numbered lists to DOCX: yes | Bulleted and numbered lists to HTML: yes |
| Simple tables | Simple tables to MD: yes | – Simple tables to TXT: no | Simple tables to DOCX: yes | Simple tables to HTML: yes |
| Merged table cells | – Merged table cells to MD: no | – Merged table cells to TXT: no | Merged table cells to DOCX: yes | Merged table cells to HTML: yes |
| Links | Links to MD: yes | – Links to TXT: no | Links to DOCX: yes | Links to HTML: yes |
| Fixed page layout | – Fixed page layout to MD: no | – Fixed page layout to TXT: no | Fixed page layout to DOCX: yes | – Fixed page layout to HTML: no |
What a PDF to Markdown converter cannot recover
Portable Document Format
PDF was created by Adobe in 1993 and became the open ISO 32000 standard in 2008. It fixes text, fonts, images and layout so a document looks the same on every screen and printer. PDFs can contain forms, signatures, links, attachments and searchable text, and the PDF/A variant is used for long-term archiving. The fixed layout makes PDFs hard to edit, and scanned PDFs contain only images unless OCR is applied. Every browser and operating system can open them.
- Compression
- –
- Transparency
- Yes
- Animation
- No
- Introduced
- 1993 · Adobe
MD
Markdown
Markdown is a lightweight markup syntax created by John Gruber in 2004. It uses plain-text symbols, such as # for headings, * for emphasis and - for lists, that stay readable as raw text and convert cleanly to HTML. It is the standard for README files, documentation, wikis, note apps and many static site generators. There are several dialects, such as CommonMark and GitHub Flavored Markdown, which differ in details like tables. Any text editor can open .md files.
- Compression
- –
- Transparency
- –
- Animation
- –
- Introduced
- 2004 · John Gruber
People turn a PDF into Markdown when the text has to live in plain-text tools, such as a notes vault, a docs repo or an AI retrieval index, where a heading must be a # rather than a bigger font.
Some things were never stored in a form that can be read back. A sidebar placed between columns can end up mid-paragraph. Equations typeset as positioned symbols come out as a line of characters, not LaTeX. Text drawn as vector outlines, common in logos and exported slides, contains no characters at all.
Tables with merged header cells or multi-paragraph cells are flattened into a simple grid, so expect to rewrite a complicated header row by hand. If you only need the words, PDF to TXT gives plain text; if your source is a Word draft, DOCX to Markdown keeps more structure, because a DOCX still knows which paragraphs are headings.
Markdown for LLM and RAG pipelines, Obsidian and Notion
RAG and LLM pipelines
Headings give a text splitter natural chunk boundaries, so each chunk keeps its section title, and Markdown spends far fewer tokens on markup than HTML.
Obsidian vaults
Obsidian keeps every note as a plain .md file in a folder on your disk, so the converted file drops straight into a vault.
Notion workspaces
Notion's Markdown import turns headings, lists and pipe tables into native blocks, cleaner than pasting from a PDF viewer.
For a retrieval pipeline, convert with Keep images off unless a vision model will read the figures, and pick the right Document language for scans so OCR does not invent characters that end up in your index. Then check the first two heading levels: a caption promoted to ## splits chunks in odd places, and one find-and-replace fixes it before indexing. Plain text also diffs cleanly in Git, so a new edition of a manual can be compared sentence by sentence.
PDF to Markdown questions
Which Markdown flavor does the PDF to MD converter write?
Headings, lists, emphasis and links follow CommonMark, and tables are written as GitHub-flavored pipe tables, because plain CommonMark has no table syntax. That renders correctly on GitHub, GitLab, Obsidian and VS Code.
Can I convert a scanned PDF to Markdown?
Yes. With Text recognition (OCR) on Auto, pages without a text layer are read first and the structure is rebuilt afterward, so a flat 300 DPI scan gives a cleaner outline than a skewed phone photo. For a searchable PDF instead of a .md file, use OCR PDF.
Why did a bold caption turn into a heading?
Headings are detected from size, weight and spacing, and a short bold line standing alone, such as a figure caption, can look exactly like a minor heading. Delete its # marks; the rest of the outline is unaffected.
How do I bring the Markdown file into Notion?
Choose Import in Notion's sidebar, pick Text & Markdown and select the file. Each # heading becomes a heading block and each pipe table a simple table.
Why is the .md file so much smaller than the PDF?
The PDF carries embedded fonts, images and drawing instructions for every page; Markdown keeps characters and a handful of symbols. A text-heavy report of several megabytes typically becomes a few hundred kilobytes of Markdown, which also fits a model's context window more easily.
Can I turn the Markdown back into a PDF after editing?
Yes. Any editor that exports can produce a PDF again, and UltKit has a Markdown to PDF converter for the round trip. The new PDF uses the renderer's fonts and spacing, not the original design.
- PDF to Text ConverterExtract the words from any PDF as a clean UTF-8 text file. Scans are read with OCR, page ranges are optional, and nothing but the text is kept.
- PDF to TXT ConverterConvert PDF to TXT online for free. Digital pages keep their exact characters, scanned pages go through OCR, and the result is one UTF-8 .txt file per PDF.
- PDF to Word ConverterConvert PDF to Word you can actually edit. Text reflows into normal paragraphs, scanned pages go through OCR in 34 languages, up to 20 files at a time.
- DOCX to Markdown ConverterConvert Word DOCX files to clean Markdown for GitHub, static site generators and docs-as-code. Images saved beside the .md in a ZIP, 20 files per batch.
- Markdown to PDFConvert .md and .markdown files to PDF with real headings, lists, tables, links and code blocks instead of raw asterisks. Bookmarks from your # headings.
- PDF to EPUB ConverterConvert a PDF into a reflowable EPUB that adjusts to Kindle, Kobo, Apple Books and phone screens. Font size becomes adjustable, images stay in.
