PDF to Markdown Converter
Turn a text-based PDF back into clean, editable Markdown. We read the real text layer and rebuild structure — then let you re-export to a polished PDF. Free, no signup, nothing stored.
Drop a PDF here, or click to choose
Your file is converted on the fly and never stored. Max 15 MB.
How the conversion actually works
Most “PDF to text” tools dump one long, structureless block. We reconstruct the document instead.
Reads the text layer, not pixels
Every character in a text-based PDF carries its position and font. We extract that layer directly — no image recognition, no re-typing, no guesswork about letters.
Infers headings from font size
We build a font-size histogram to find the body text size, then promote larger runs to #, ##, and ### relative to it. Your title and sections come back as real headings.
Rebuilds paragraphs & lists
Lines are grouped by position, wrapped lines are merged (and hyphenated words rejoined), and bullet/number prefixes become proper Markdown lists.
Strips repeating clutter
Running headers, footers, and page numbers that repeat across pages are detected and removed, so they don't pollute your Markdown.
What gets converted — honestly
We'd rather flag a hard case than hand you a confident wrong answer.
| Element | What happens | Support |
|---|---|---|
| Headings (title, sections) | Mapped to #, ##, ### by relative font size | ✓ |
| Paragraphs | Wrapped lines merged; hyphenated breaks rejoined | ✓ |
| Bulleted & numbered lists | Detected from markers → - and 1. | ✓ |
| Bold & italic | Detected from font weight/style → ** and * | ✓ |
| Code blocks | Monospace runs fenced with ``` | ✓ |
| Page breaks | Preserved as <!-- pagebreak --> between pages | ✓ |
| Running headers / footers / page numbers | Detected and removed automatically | ✓ |
| Complex tables | Flagged for a quick manual tidy (no false guesses) | ~ |
| Images & figures | Not extracted (text-only conversion) | ✗ |
| Scanned / image-only PDFs | No text layer — needs OCR first | ✗ |
PDF to Markdown vs other methods
| Method | Structure kept | Effort | Privacy |
|---|---|---|---|
| mdtopdf.io (this tool) | Rebuilds headings, lists, code, paragraphs | One upload | Nothing stored |
| Copy & paste from a PDF viewer | Loses all structure; broken line wraps | Heavy manual cleanup | Local |
| Pandoc | No native PDF input — needs extra tooling | Install + CLI | Local |
| Generic online OCR converters | Re-recognizes pixels; error-prone on real text | Upload, often watermarked | Often stored |
When people reach for this
Re-edit an old PDF
You only have the exported PDF and need to change it. Convert to Markdown, edit the text, and re-export a clean PDF.
Feed clean text to an LLM
Paste a PDF's contents into ChatGPT or Claude without broken line-wraps and page-number noise dragging down the model.
Archive docs as plain text
Keep documentation in version control as Markdown instead of opaque binary PDFs, so diffs and search actually work.
Move notes between tools
Pull a PDF handout into Obsidian, Notion, or any Markdown editor with its headings and lists intact.
Need the other direction?
Have Markdown and want a PDF? Our Markdown to PDF converter renders GitHub-flavored Markdown into a styled, paginated document — with the same privacy model.
PDF to Markdown FAQ
How does this convert a PDF to Markdown?＋
It reads the PDF's embedded text layer — the actual character, position, and font data inside the file — not a picture of the page. From the geometry it infers structure: larger fonts become headings, bullet and number prefixes become lists, monospace runs become code blocks, and wrapped lines are merged back into paragraphs. The result is clean Markdown, not a wall of unstructured text.
Is it free and do I need an account?＋
It is completely free with no signup. Upload a PDF up to 15 MB and get Markdown back immediately.
Are my PDFs stored or used for training?＋
No. The file is streamed to a short-lived worker, parsed in memory, and discarded. Nothing is written to disk, indexed, or used to train anything.
Can it convert a scanned PDF or an image-only PDF?＋
No. If a PDF is a scan (a photo of a page) it has no text layer, so there is nothing to extract. You would need OCR software first to add a text layer, then convert here. The tool tells you when a file appears to be image-only.
What happens to tables?＋
Simple text is reconstructed reliably. Complex multi-column tables are the hardest part of any PDF, because PDFs store table cells as loose positioned text with no row/column metadata. When a tabular layout is detected the tool flags it so you can quickly tidy those sections by hand rather than trusting a wrong guess.
Does it keep bold, italics, and headings?＋
Yes. Bold and italic runs are detected from the font name and wrapped in ** and * markers. Heading levels are assigned by relative font size, so a document's title, sections, and sub-sections map to #, ##, and ###.
Can I edit the result and turn it back into a PDF?＋
Yes — that round trip is the point. Click ‘Open in PDF editor’ to send the converted Markdown straight into our editor, tweak it, pick a template, and export a fresh, well-formatted PDF.
Which PDFs work best?＋
Text-based PDFs exported from tools like Word, Google Docs, LaTeX, Typst, or our own Markdown-to-PDF converter convert most cleanly, because their text layer is well structured. Heavily designed marketing PDFs with overlapping layers convert less cleanly.