How It Works Under the Hood
Loads your PDF byte buffer into a dedicated `pdfjs-dist` Web Worker, parses the page selector expression (`all` or ranges like `1-4, 7`), reads text items with baseline `(x, y)` coordinates, groups items along shared horizontal lines (`yTolerance = 3.5`), sorts left-to-right, and formats clean Plain Text or Markdown output.
Step-by-Step Instructions
Choose your PDF file (processed strictly in browser memory).
Optionally enter a page range (`all` or `1-3, 5`) and click Re-Extract.
Pick `Plain Text (.txt)` or `Markdown (.md)` and toggle Page Headers if desired.
Copy extracted text to clipboard or export directly to `.txt` / `.md`.
Key Features & Performance
- Line-baseline geometry reconstruction prevents scrambled multi-word column fragments
- Flexible page selector supports single pages and ranges (`1-3, 6, 9-12`)
- Exports directly to Plain Text (`.txt`) or Markdown (`## Page N` sections)
- Reports extracted page count and total word count immediately
- Zero server transfer: built for legal agreements, medical PDFs, and financial filings
Methodology, Formulas & Technical Standards
Text tokens are clustered by baseline `y` proximity within `3.5` PDF user units and sorted horizontally by `transform[4]` (`x`) before collapsing internal whitespace cleanly.
Common Real-World Use Cases
- Pulling clauses out of NDAs or vendor contracts to analyze or summarize locally
- Extracting specific chapters from multi-hundred-page technical PDF manuals
- Converting searchable invoice or receipt PDFs into text for bookkeeping paste-in
- Preparing research excerpts in Markdown format for note apps like Obsidian or Notion
Worked Examples & Edge Cases
Selective Page Range Extraction
Pages: `1-2, 5` on a 12-page contract
Extracts only pages 1, 2, and 5 with clear page dividers
Avoids copying boilerplate appendices when you only need key clauses.
Scanned PDF Detection Note
Image-only scanned document page
_No extractable text layer on this page._
If a PDF lacks an embedded text layer, switch to our Batch OCR tool.
Limitations, Scope & Accuracy Notes
- • Only extracts embedded digital PDF text layers — scanned image-only PDFs without OCR yield empty text (use Batch OCR for image scans).
- • Complex multi-column academic layouts may merge adjacent columns on the same baseline if vertical spacing aligns tightly.
Local Browser Processing & Privacy
Traditional web converters upload your documents, snippets, or media files to remote servers. Local PDF Text & Page Extractor runs inside your browser tab:
Your input text and files are processed locally without uploading payloads to a processing server.
Microdeck does not log or store the files or inputs you paste into tool workspaces (normal static hosting/CDN requests still load page assets).
Working outputs stay in your browser tab and are released when you reset the tool or close the tab.
Frequently Asked Questions
Why did a scanned PDF extract blank text?
Scanned PDFs store flat pictures of pages rather than selectable character glyphs. Use Microdeck's offline Batch OCR tool to recognize scanned PDF pages.
Does this upload my confidential PDF to Cloudflare or external APIs?
No. `pdfjs-dist` parses bytes inside an isolated local browser worker.
Can I extract just specific pages instead of the whole PDF?
Yes. Enter expressions like `1-3, 5` in the Pages input and click Re-Extract.
What is the difference between `.txt` and `.md` export?
`.txt` uses `=== Page N ===` banners, while `.md` adds Markdown `## Page N` headers and blank-page placeholders.
Explore Related Micro-Tools
PDF Metadata Inspector & Scrubber
View hidden Author, Creator, Producer & XMP metadata inside PDFs — edit or scrub cleanly offline
Image & Scanned PDF to Text Extractor
Extract editable text from scanned PDFs and photos privately right inside your browser
Word Counter, Grapheme & Readability Analyzer
Count Intl.Segmenter graphemes, UTF-8 bytes, speaking/reading times, top keywords & Flesch-Kincaid grade