How to Extract Text From a PDF (Three Methods)

Not all PDFs are the same. A PDF created by exporting from Word contains actual text objects you can select and copy. A PDF created by scanning a paper document contains images of text. The methods for extracting text from each are different.

Method 1: Native PDF Text (The Easy Case)

Open in any PDF viewer, select all text (Ctrl+A or Cmd+A), copy and paste. If this works, you're done. Most PDFs created from digital sources (Word exports, InDesign exports, HTML-to-PDF) are native and allow text selection. If the text selects but looks garbled when pasted, the PDF may have custom font encoding — in that case, try a specialized extractor.

Method 2: Scanned PDFs — OCR Required

If clicking in the PDF selects page regions rather than individual words, it's image-based. Google Docs is the quickest free solution: upload to Google Drive, right-click > Open with > Google Docs. It runs OCR and creates an editable document with the recognized text. Quality varies by scan quality.

Method 3: Batch Extraction for Developers

PyMuPDF (Python) is fast and handles most PDFs: import fitz; doc = fitz.open('file.pdf'); for page in doc: text += page.get_text(). For scanned PDFs, combine with Tesseract: pdf2image converts pages to images, then pytesseract runs OCR. For larger scale extraction, Adobe's PDF Extract API handles both native and scanned PDFs with high accuracy.

When Text Looks Wrong After Extraction

Column-based documents (newspapers, academic papers) extract text in reading order incorrectly — it mixes columns rather than reading each column separately. Two-column PDFs are a known challenge; specialized tools that understand layout (PDFPlumber's `page.extract_text(x_tolerance=5)`) do better here. Tables similarly require specialized extraction to maintain structure.