Extract text from PDF into clean Markdown
Drop a PDF and get its text back as readable Markdown – not a jumbled copy-paste. Works on digital PDFs and on scanned, image-only files through built-in OCR, and keeps headings, lists, real tables, formulas and links. Free in the browser, no sign-up.
Yes – any PDF, digital or scanned
Open the converter, add your PDF as a file or a direct URL, and the text comes back as Markdown. Markdown is plain text with lightweight structure, so you can copy it straight into a document, an editor or a chat, or download it as a .md file and open it anywhere. If the PDF is a scan, OCR runs automatically, so pages that would normally return nothing on copy-paste still produce real, selectable text. This is a PDF text extractor, not a spreadsheet builder: the result is readable text, not JSON or Excel.
How to extract text from a PDF in three steps
No account, no install, no separate OCR tool.
Open the converter
Use the web app in any browser, or install the Chrome extension to extract text from PDFs you find while browsing. Both work anonymously.
Add the PDF
Drag in the file, choose it from disk, or paste a direct PDF URL. Digital PDFs are read from their text layer; image-only pages are OCR'd automatically, and you can toggle force OCR when the embedded text layer is wrong.
Copy or download the text
The job goes queued, processing, ready. Read the extracted text in the rendered or raw preview, then copy it to your clipboard or download a .md file – plain text that opens in any editor.
Extracting text from many PDFs? The same pipeline is a REST API and hosted MCP, so scripts and agents can pull the text without touching the UI.
Digital PDF versus scanned PDF
Whether a PDF has extractable text at all depends on how it was made. The converter handles both cases, but they behave differently.
Digital PDF (has a text layer)
Scanned or image-only PDF
Raw extracted text versus structured Markdown
A PDF stores glyphs at coordinates, not sentences. Naive extraction reads them roughly in drawing order, so a two-column page interleaves, table rows scatter, headings look like body text and formulas fall apart. Here is the same table extracted both ways.
Naive raw extraction
Quarter Revenue Growth Q1 $1.2M +8% Q2 $1.4M +17% Q3 $1.9M +36%
Readable by nobody: rows and columns are lost, so a number can no longer be matched to its heading.
Structured Markdown
| Quarter | Revenue | Growth | | ------- | ------- | ------ | | Q1 | $1.2M | +8% | | Q2 | $1.4M | +17% | | Q3 | $1.9M | +36% |
Still plain text, still copy-pasteable anywhere – but the facts stay attached to their headers.
On "PDF data extraction": people often mean pulling named fields into a spreadsheet or JSON. That is not what happens here. This tool extracts the document's text and structure as Markdown; if you need structured records, parse those Markdown tables downstream yourself.
Headings, tables, formulas, links and reading order
Recognizing characters is the easy half. These are the things a plain text dump usually loses and the converter tries to keep.
Headings & lists
Section titles come back as Markdown headings and bullets stay bullets, so the document keeps its outline instead of becoming one long paragraph.
Real tables
Columns are rebuilt as genuine Markdown tables rather than misaligned lines. See extracting tables from PDF.
Formulas
Mathematical notation is preserved instead of being flattened into garbled characters. More on formulas and LaTeX.
Links & footnotes
Where the PDF contains them, hyperlinks and footnotes carry over as Markdown links instead of being dropped.
Reading order
Multi-column pages are read column by column, so sentences do not interleave the way a raw dump makes them.
Scanned pages
OCR turns image-only pages into selectable text in the same pass, with a selectable OCR language.
Where text extraction is still hard
No extractor is perfect, and this one makes no guarantee of flawless OCR or flawless reading order on every document. These are the cases that most often need a second look.
Poor scans
Faint, skewed or low-resolution pages, phone photos and copies of copies reduce OCR accuracy. Rescanning straight at a higher resolution is the quickest fix.
Exotic layouts
Magazine grids, nested or merged table cells, sidebars and rotated pages can still come out in an unexpected order.
Bad text layers
Some PDFs embed a wrong or partial text layer. Turn on force OCR so the page images are re-read instead of that layer.
Handwriting
The OCR targets printed and typeset text. Handwritten notes are not reliably recognized.
Very long documents
Each conversion runs up to a per-tier time budget. A long file is returned partially and flagged truncated rather than failing outright.
Engine choice helps
If a page comes out badly, try the other engine: MinerU is robust on scans and complex layouts, Docling is fast on clean, simple pages.
Extract text without an account
Free tier limits
Paid tiers raise every limit and add a longer time budget for heavy scans. Compare plans →
Your document stays yours
Extracting a particular kind of PDF
Scanned documents
Image-only files, force OCR, scan quality and what to do when nothing is recognized.
Scanned PDF to Markdown →Tables
Get aligned Markdown tables out of reports and statements instead of scattered numbers.
PDF tables to Markdown →Invoices & statements
Line items and totals as readable Markdown you can review or process further.
Invoice PDF to Markdown →Ready to go? Convert a PDF to Markdown now, or ask us anything on the support page.
Common questions
How do I extract text from a PDF?
Open the web app or the Chrome extension, add the PDF as a file or a direct URL, and wait for the job to finish. The extracted text appears as Markdown you can copy to the clipboard or download as a .md file. No account and no install are required.
Can I extract text from a scanned PDF?
Yes. A scanned PDF holds pictures of pages rather than text, so plain copy-paste returns nothing. Image-only pages are OCR'd automatically into selectable text, and you can force OCR when a PDF has a bad or partial text layer. See scanned PDF to Markdown.
What format is the extracted text?
Markdown. It is plain text with lightweight structure, so headings, lists, tables and links survive as readable text rather than formatting codes. You can copy it or download a .md file and open it in any text editor. There is no separate .txt download.
Why is copy-paste from a PDF so messy?
A PDF stores glyphs at coordinates, not sentences. Naive extraction reads them roughly in drawing order, so columns interleave, headings look like body text, table rows scatter across lines, formulas break apart and hyphenated words split. Rebuilding the reading order and the structure is the part that makes the text usable.
Does it extract tables and formulas too?
Yes. Columns are rebuilt as real Markdown tables instead of scattered lines, and mathematical notation is preserved rather than flattened into garbled characters. Hyperlinks and footnotes carry over as Markdown links where the PDF contains them.
Is there a free PDF text extractor with no sign-up?
Yes. You can extract text anonymously in the browser on the free tier: 3 active slots, 10 MB files, a 15-minute time budget per document and 1-hour retention. Paid plans raise every limit – see pricing.
Can it extract data into JSON, CSV or Excel?
No. The output is Markdown only. The converter does not export JSON, CSV or spreadsheets and does not pull named fields or key-value pairs out of a document. If you need structured data, parse the Markdown tables yourself downstream.
Are my files private?
Files and results auto-delete after the retention window, which is 1 hour on the free tier, and you can delete any job manually. Document content is never used for advertising or to train models.