Text extraction from PDF is what makes copy-paste and full-text search possible. This tool uses pdf.js to read every text stream, reconstruct paragraph boundaries where the layout permits, and give you a clean plain-text output. Perfect for feeding PDFs into LLMs, populating search indexes, or just grabbing quotes.
Paragraph reconstruction
Naive text extraction from PDFs produces one line per glyph run, which is unusable. The tool reassembles lines into paragraphs based on line spacing, indentation, and hyphen removal. Works well on single-column documents; multi-column layouts (papers, magazines) may need manual cleanup.
Scanned PDFs
Scanned PDFs contain images, not text — extraction produces nothing usable. Run the OCR PDF tool first to make the file searchable, then extract text. The tool warns when a PDF appears to be scan-only.
Output options
Copy to clipboard, download as .txt, or download with page markers (--- Page N --- separators). Toggle Unicode NFC normalization for cleaner output when the source uses combining diacritics.