PDF to Text & OCR - DOCX, Scanned PDF, Images to Text
Pricing
from $3.00 / 1,000 documents
Convert PDF, DOCX, PNG, JPG and TIFF files to clean text, page by page. Text-layer PDFs are read directly without OCR; scanned pages and images are OCR'd in 7 languages. Optional Markdown and RAG chunks. $0.003 per document, $0.006 per OCR page, failed files free.
Direct links to PDF, DOCX, PNG, JPG, TIFF, BMP, GIF or WEBP files (http/https), one per line. The file type is detected from the content, not the extension.
Keys of files you uploaded to a key-value store (upload each file with its real Content-Type, e.g. application/pdf). Combine with 'URLs' freely.
Name of the store that holds the keys above. Leave empty to use this run's default store.
Small files passed directly in the input as {"name.pdf": ""}. Total input size is limited by Apify (about 9 MB); use URLs or a key-value store for anything bigger.
Languages of scanned pages and images. English is always added. Choose every language that occurs in your documents; each extra language slightly slows OCR.
[ "eng"]auto: OCR only pages that have no text layer (default, cheapest). force: OCR every page (use when a PDF has a broken or garbage text layer). never: text layers only, no OCR.
Add a 'markdown' field. DOCX headings, lists and tables are preserved; PDF and OCR text becomes paragraphs with page markers.
0 = no chunks. Otherwise adds a 'chunks' array of about this many characters, cut on paragraph boundaries, each with its page range.
How much of the previous chunk is repeated at the start of the next one (paragraph granularity).
Pages after this number are not processed (and not charged). The output tells you when a document was truncated.
Stop after this many documents (the rest are skipped).
Larger files are reported as failed (free).
Resolution used to rasterize PDF pages before OCR. 300 is the best default; 200 is faster and fine for clean scans.