How do I convert a PDF to Markdown?

Upload your PDF, choose a Markdown flavor (GitHub, Obsidian, etc.) and a mode, then click Convert. The text and structure are extracted and rebuilt as Markdown, which you can preview, copy or download as a .md file — all on your device.

How do I convert PDF to MD for free?

Drop your PDF in, keep the default GitHub Flavored Markdown, and Convert. It is completely free with no sign-up, no watermark and no daily limit; a developer API is available for automation.

Is this PDF to Markdown converter free?

Yes. Converting PDFs to Markdown is free with no account, watermark or per-file cap. The browser conversion is free forever; a REST API offers automated and high-volume conversion on paid tiers.

Are my PDFs uploaded to a server?

No. The entire conversion runs in your browser using pdf.js — your PDF is read locally, converted to Markdown on your device, and offered back as a download. Nothing is transmitted, which makes it private and GDPR-friendly by design.

Does it convert PDF to GitHub Markdown?

Yes. Choose the GitHub (GFM) flavor for GitHub Flavored Markdown — pipe tables, task lists, fenced code blocks and alerts — ready to paste into a repository, README, issue or wiki.

Does it convert PDF to Obsidian Markdown?

Yes. The Obsidian flavor outputs YAML frontmatter and Obsidian-friendly structure so your converted notes drop straight into an Obsidian vault. Use Knowledge Base mode for the best note structure.

Does it convert PDF to Notion Markdown?

Yes. The Notion flavor produces Markdown that imports cleanly into Notion (headings, lists, tables, code and quotes), so you can turn a PDF into a Notion page in seconds.

Does it support MkDocs, Docusaurus, Hugo and Jekyll?

Yes. Each has a dedicated flavor that adds the right YAML frontmatter (title, date, tags) and admonition-friendly structure for MkDocs, Docusaurus, Hugo and Jekyll static-site generators.

Can it create a README from a PDF?

Yes. The README flavor formats the document as a project README — a top H1 title, a summary blockquote and clean sections — ideal for turning a spec or doc into a repository README.md.

Does it detect headings automatically?

Yes. Headings are detected from font sizes (and bold short lines), then mapped to H1–H6 by size rank, so your Markdown keeps the document’s outline. You can toggle heading detection off if you prefer flat text.

Does it preserve tables?

Yes. Tables are detected by clustering the horizontal positions of text into columns and rebuilt as Markdown (GitHub) tables with a header row and alignment. Complex or borderless tables are best-effort; you can disable table detection if needed.

Can I export tables as CSV or JSON?

The converted tables appear as Markdown tables in the output, and the full document structure (including table rows) is available as the Structure JSON export in Developer view — which you can transform to CSV/JSON downstream. Dedicated CSV/JSON table export is on the API.

Does it detect code blocks?

Yes. Runs of monospaced text are detected as code and wrapped in fenced code blocks, with an automatic language guess (JavaScript, Python, C/C++, Java, SQL, JSON, YAML, XML, Bash) for syntax highlighting.

Does it detect lists and checklists?

Yes. Bulleted lists, numbered lists and checkboxes ([ ] / [x]) are detected and converted to proper Markdown list syntax, with nesting based on indentation.

Does it handle math and equations?

Lines that use math fonts or contain mathematical symbols are detected and wrapped as KaTeX/MathJax-compatible LaTeX blocks ($$ … $$), preserving the content. Full equation OCR to precise LaTeX (Mathpix-style) is available via the API.

Does it extract images?

Yes. Embedded images are located from the PDF’s drawing operators and extracted. Choose to embed them as base64 in the Markdown, save them as separate PNG or WebP files (referenced with Markdown image links), or skip images entirely.

Can it embed images as base64?

Yes. Set image handling to “Embed (base64)” and every image is inlined as a data URI directly in the Markdown, producing a single self-contained .md file with no external assets.

Can it convert images to WebP?

Yes. Choose “Extract as WebP” to save extracted images in the efficient WebP format inside the downloaded ZIP, with Markdown links pointing to them.

Does it preserve hyperlinks?

Yes. Link annotations in the PDF are read and mapped to the text they cover, producing Markdown links [text](url). Internal cross-references and citations are preserved as text.

Can it OCR scanned PDFs?

Yes. The tool detects scanned/image-only PDFs and, when OCR is enabled, recognises the page images into text with tesseract.js before generating Markdown. OCR supports English, Hindi, Arabic, French, German, Spanish, Chinese, Japanese, Korean and more.

Which OCR languages are supported?

English, Hindi, Arabic, French, German, Spanish, Chinese (Simplified), Japanese, Korean, Russian, Portuguese and Italian — powered by tesseract.js, which loads on demand only when OCR is enabled.

How do I convert a scanned PDF to Markdown?

Turn on OCR in settings, pick the document’s language, then convert. Each page image is recognised into text and assembled into Markdown. Scanned conversions take longer and lose some layout fidelity compared with text PDFs.

Does it work for research papers?

Yes. Research-paper PDFs convert well — headings, sections, references and links are reconstructed, and AI insights extract keywords, an extractive summary and entities. Use High Accuracy mode and enable math detection for equation-heavy papers.

Can I convert a book or large PDF?

Yes. The engine handles long documents — books, textbooks and 1000+ page manuals — by streaming page extraction. Use the page-range option to convert a portion, and a desktop browser for the very largest files.

Can I convert only certain pages?

Yes. Enter a page range like “1-10, 12, 20-25” and only those pages are extracted and converted.

What is AI Reconstruction mode?

AI Reconstruction mode applies the full heuristic structure-detection and cleanup pipeline (fix hyphenation, remove duplicates, repair paragraph breaks, normalise) and an RAG-optimised layout. It is on-device and rule-based — no content is sent to an external LLM.

What does the AI cleanup do?

It joins words split by end-of-line hyphenation, removes duplicated running text, repairs paragraph breaks split across lines, strips repeated headers/footers and page numbers, and normalises whitespace and encoding — producing clean, readable Markdown.

What are the optimization profiles?

Readable keeps human-friendly spacing; Compact minimises whitespace; SEO produces clean heading structure; RAG-optimized inserts chunk separators on top-level headings for retrieval systems; LLM-friendly is token-efficient and converts images to alt-text references.

What is RAG-optimized Markdown?

RAG (Retrieval-Augmented Generation) optimised output splits the document into self-contained chunks at top-level headings, separated by horizontal rules, so each section can be embedded and retrieved independently by a vector database — ideal for building AI knowledge bases.

Can I generate LLM-friendly / AI training Markdown?

Yes. The LLM-friendly optimisation produces compact, structure-preserving Markdown that is token-efficient for AI training and prompting, with images reduced to alt-text references so the text stays clean.

Does it generate keywords and a summary?

Yes. The AI insights panel extracts the top keywords and topics (TF ranking), an extractive summary (top sentences by keyword relevance), named entities (people, organisations, dates, emails, URLs), a complexity score and reading time — all computed locally.

Is the summary generated by an LLM?

No. The summary is extractive and heuristic — it selects the most relevant existing sentences using keyword scoring, computed entirely on your device. It is honest, deterministic and private, not an LLM-generated paraphrase.

Does it extract entities?

Yes. The entity extractor finds emails, URLs, dates, numbers and Title-Case people/organisation names using pattern and capitalisation heuristics, grouped and ranked by frequency — useful for indexing and knowledge graphs.

What is the document complexity score?

It is derived from the Flesch Reading Ease formula (sentence length and syllables per word). A higher complexity score means denser, harder-to-read text; it helps you gauge how technical a document is.

Does it have a live preview?

Yes. The result viewer offers a rendered Markdown Preview, a raw Source view with in-output search, a Split view (source + preview side by side), an AI insights tab and a Developer tab — so you can verify the output instantly.

Can I search within the output?

Yes. The Source view has a search box that highlights matches in the generated Markdown, so you can quickly find content in long documents.

What does Developer mode show?

Developer mode exposes the raw Structure JSON (the parsed block model), the Metadata JSON (document info, fonts, stats), a Markdown linting report and Markdown statistics — ideal for pipelines and debugging.

Does it lint the Markdown?

Yes. The built-in linter checks for missing or multiple H1s, heading-level jumps, empty links, trailing whitespace and inconsistent table columns, and reports issues with line references.

Can I download the images separately?

Yes. Download the result as a ZIP, which includes the Markdown file, an images/ folder with the extracted images, and the structure and metadata JSON exports.

Can I copy the Markdown to the clipboard?

Yes. The Copy button copies the full Markdown output to your clipboard in one click, ready to paste into GitHub, Obsidian, Notion or your editor.

Can I convert multiple PDFs at once?

Yes. Add several PDFs (or a folder) and they are queued and converted one after another, each with its own progress, preview and download. Grab them all as a single ZIP when done.

How accurate is the conversion?

For text-based PDFs, structure reconstruction is high quality — headings, paragraphs, lists, code and most tables come through cleanly. Accuracy depends on the PDF’s internal structure; complex multi-column or graphic-heavy layouts are simplified. Each conversion shows an estimated accuracy score.

Why is my multi-column PDF jumbled?

Multi-column layouts can interleave text because PDF stores text by position, not reading order. Use High Accuracy mode; for heavy multi-column academic PDFs, the API’s advanced layout analysis handles columns better.

How long does conversion take?

Most PDFs convert in a few seconds. Large documents take longer because each page is parsed, and OCR is slower still (each page is recognised). A live dashboard shows the current stage and ETA.

Is there a file-size limit?

Because conversion runs in your browser, the limit is your device’s memory rather than a server cap. Large PDFs convert comfortably on desktop; very large or scanned files are slower on phones.

Does it work offline?

After the page loads once, text-PDF conversion works offline since everything runs on your device. OCR downloads its language model on first use, so the first OCR needs a connection.

Which browsers are supported?

Any modern browser — Chrome, Edge, Firefox, Safari, Brave and Opera — on desktop and mobile.

Does it work on phones and tablets?

Yes. The converter is fully responsive and runs in mobile browsers on Android and iOS; large or scanned PDFs are slower on phones due to memory limits.

How does this compare to Mathpix?

Mathpix specialises in converting math and scientific PDFs to LaTeX/Markdown with high equation fidelity. This tool covers general PDF-to-Markdown (headings, tables, code, lists, images) in your browser for free, with heuristic math detection; precise equation OCR is offered via the API.

How does this compare to Pandoc?

Pandoc is a powerful document converter, but it doesn’t natively read PDFs well (PDFs aren’t structured documents). This tool reconstructs structure directly from the PDF layout in your browser and outputs Pandoc-compatible Markdown you can post-process with Pandoc if you like.

PDF to Markdown Converter

Extract text, tables, code, images and structure from any PDF into clean Markdown — GitHub, Obsidian, Notion and more. With OCR, AI insights and RAG export, all privately in your browser. No watermark, no sign-up.

10 Markdown flavors100% privateOCR + AI insights

PDF → clean structured Markdown100% private — runs in your browserOCR · tables · code · AI insights

Drag & drop your PDF

or browse, upload a folder, or import from a URL — converts to Markdown

Text & scanned PDFs · books · research papers · technical manuals · 1000+ pages

Your queue is empty. Add a PDF above — text and scanned PDFs are converted to clean Markdown, privately on your device.

Why this is the best free PDF to Markdown converter

Real structure detection

Headings, paragraphs, lists, checklists, code blocks, tables, blockquotes, links and images rebuilt as clean Markdown.

100% private

Parsed with pdf.js in your browser. Your PDF is never uploaded; OCR runs on-device too.

10 Markdown flavors

Standard, GitHub, Obsidian, Notion, MkDocs, Docusaurus, Hugo, Jekyll, README and Knowledge Base.

OCR for scanned PDFs

Detects image-only PDFs and recognises them in 12 languages with tesseract.js for searchable Markdown.

AI insights & RAG

On-device keywords, extractive summary, entities and a complexity score, plus RAG/LLM-friendly output.

Developer mode

Live preview & split view, in-output search, structure & metadata JSON exports and a Markdown linter.

What is PDF to Markdown conversion?

PDF to Markdown conversion turns a PDF document into clean, structured Markdown — extracting the text and rebuilding the document’s headings, paragraphs, lists, tables, code blocks, blockquotes, links and images using Markdown syntax. The result is a lightweight, editable, version-controllable text file you can use in documentation, notes, GitHub and AI pipelines.

PDFs store text by position on the page rather than as a logical document, so good conversion is really structure reconstruction: reading every text run with its font size and coordinates, inferring headings from larger fonts, lists from bullet markers, code from monospaced fonts, and tables from column alignment, then serialising it all to Markdown. This tool does that entirely in your browser with pdf.js, and OCRs scanned PDFs on request.

Why convert PDF to Markdown?

Markdown turns a locked-down PDF into clean, reusable text. The most common reasons people convert:

Documentation — turn specs and manuals into docs-site content (MkDocs, Docusaurus, Hugo, Jekyll).
Knowledge bases — import PDFs into Notion, Obsidian and wikis as proper notes.
GitHub publishing — paste clean GFM into READMEs, issues and repository wikis.
AI training datasets — produce clean, structured text for fine-tuning and prompting.
RAG systems — chunk documents for embedding into vector databases.
Editing & version control — Markdown gives perfect Git diffs and edits anywhere.

Convert a PDF to Markdown in 3 steps

Upload your PDF

Drag in a PDF (or import from a URL). The tool instantly analyses it — pages, words, images, tables, links, fonts, metadata, language and reading time — and shows page thumbnails and the document structure, all in your browser.

Choose a Markdown format

Pick an output flavor (Standard, GitHub, Obsidian, Notion, MkDocs, Docusaurus, Jekyll, Hugo, README or Knowledge Base) and a conversion mode. Enable OCR for scanned PDFs and tune structure, image and cleanup options.

Convert & download

Press Convert. Watch the live progress dashboard, then preview the Markdown side-by-side, copy it, download the .md (or a ZIP with images and JSON exports), and review AI insights and the linter report.

PDF vs Markdown

PDF is a fixed-layout format for finished documents; Markdown is a lightweight plain-text format for editable, structured content. The table below compares the two where it matters.

Feature	PDF	Markdown
Editing	Hard to edit; needs special tools	Plain text; edit anywhere
Portability	Fixed layout, large binary	Tiny, universal text file
Version control	Binary — poor Git diffs	Perfect line-by-line Git diffs
Searchability	OK if text-based, hard if scanned	Instantly grep-able plain text
AI / RAG compatibility	Needs parsing first	Clean, chunkable, embedding-ready
Knowledge management	Static document	Native to Obsidian, Notion, wikis
Publishing	Download/print	Powers GitHub, docs sites & blogs

Supported output formats

Convert your PDF into the exact Markdown flavor your workflow needs.

PDFMarkdown (.md)

PDFGitHub Markdown

PDFObsidian Markdown

PDFNotion Markdown

PDFMkDocs Markdown

PDFDocusaurus Markdown

PDFHugo / Jekyll

PDFREADME Markdown

PDFKnowledge Base

Best PDF to Markdown settings

Different documents call for different settings. These recommendations get a great result on the first try.

Technical documentation

Documentation mode · GitHub or MkDocs flavor · keep code & tables · SEO optimisation

Research papers

High Accuracy mode · math detection on · keywords + summary + entities · GitHub flavor

Books & long PDFs

High Accuracy · strip headers/footers · fix hyphenation · page range for sections

Wiki / Obsidian notes

Knowledge Base mode · Obsidian flavor · frontmatter on · readable optimisation

AI knowledge base / RAG

AI Reconstruction mode · RAG-optimized profile · embed or skip images

Scanned documents

Enable OCR · select language · High Accuracy · normalize whitespace

Quick text grab

Fast mode · skip images · Compact optimisation

AI-ready & RAG-optimized output

Markdown is the preferred format for feeding documents to AI systems: its clean, predictable structure (headings, lists, tables, code) chunks and embeds well. This converter includes an RAG-optimized profile that splits the document into self-contained sections at top-level headings, and an LLM-friendly profile that keeps the text token-efficient.

On-device AI insights add value without sending your content anywhere: TF keyword and topic extraction, an extractive summary, entity recognition (people, organisations, dates, emails, URLs), a complexity score and reading time — all computed locally and honestly labelled as heuristic, not LLM-generated.

Privacy, security & data handling

Your documents stay yours. The conversion runs entirely in your browser — your PDF is read, parsed and converted on your device, and nothing is uploaded or logged. OCR downloads its language model locally and recognises pages on-device.

There is nothing to store on our side, and files are released the moment you close the tab — supporting GDPR-compliant handling of sensitive documents by design. For automated, server-side conversion at scale, the developer API offers the same engine with batch jobs, webhooks and team controls.

PDF to Markdown API

Automate conversion with the PDF to Markdown API: PDF-to-Markdown, OCR, AI structure detection, batch and asynchronous jobs, webhooks, API keys, team accounts and usage analytics — backed by the same engine as the web tool plus advanced server-side layout analysis.

Pricing scales from a Free tier for prototyping to Pro, Business and Enterprise plans that add precise math/LaTeX OCR, advanced multi-column layout analysis, vector-database chunk exports and SLA-backed support.

Free

$0/mo

100 conversions / mo
All 10 Markdown flavors
Structure & metadata JSON
Community support

Start free

Popular

Pro

$19/mo

10,000 conversions / mo
OCR + math/LaTeX
Async jobs & webhooks
RAG chunk export

Go Pro

Business

$69/mo

100,000 conversions / mo
Advanced layout analysis
Team accounts & API keys
Usage analytics

Choose Business

Enterprise

Custom

Unlimited volume & SLA
Vector DB pipelines
Content migration suite
Dedicated support

Contact sales

Frequently Asked Questions

A PDF to Markdown converter turns a PDF document into clean, structured Markdown (.md) — extracting the text and reconstructing headings, lists, tables, code blocks, blockquotes, links and images as proper Markdown syntax. This tool parses the PDF with pdf.js in your browser, detects structure from font sizes and layout, and serialises it to Markdown. Nothing is uploaded.

Troubleshooting guide