PDF Text Extractor & OCR: PDF and Scanned PDF to Text

Give it links to PDF files and get back each document's text, ready for search, analysis or an AI pipeline. Scanned PDFs, which are pictures of pages, can be read with OCR.

You pay only for PDFs whose text was extracted. Links that are broken, are not PDFs, or are password-protected come back with a status and cost nothing. OCR is off unless you switch it on, and it is charged only for the pages that actually need it.

What you get for each PDF

Field	What it is
text	The full text, pages separated by a blank line, with words that were hyphenated across line breaks rejoined
pages	Optional: the text of each page separately, with its page number and whether it came from stored text or from ocr
pageCount, wordCount, characterCount	Size of the document
title, author, subject, creator, producer	The document's own metadata
createdAt, modifiedAt	Document dates in standard format
looksScanned	true when nearly every page had no stored text, which means the PDF is images of text
ocrPageCount, ocrConfidence	How many pages were read with OCR, and the OCR engine's own confidence from 0 to 100
ocrIncomplete	true when some scanned pages were left unread because of your page limit or spending limit
status	ok, download-failed, not-a-pdf, unreadable or invalid-input

Scanned PDFs and OCR

Turn on Read scanned pages with OCR and every page without stored text is drawn at 300 dots per inch and read with Tesseract, an open-source OCR engine. Pages that already have text are left alone and are never charged for OCR, so a PDF that mixes typed and scanned pages costs only for the scanned ones.

• Languages: English, German, French, Spanish, Italian, Portuguese, Dutch, Swedish, Danish, Norwegian, Finnish and Polish. Pick the language of the document for the best result.
• Tables: each table row comes out as one line of text, in reading order.
• What to expect: clean printed text is read reliably. In our own test page, ordinary typefaces were read without error, while an old-style typeface whose digits vary in height had a 9 misread as 0. If figures matter, check them, and treat a low ocrConfidence as a warning.
• Not suitable for: handwriting, and photos of pages taken at an angle or in poor light.

Input

{
  "pdfUrls": ["https://arxiv.org/pdf/1706.03762"],
  "includePages": true,
  "ocrScanned": true,
  "ocrLanguage": "eng",
  "maxOcrPages": 50
}

maxOcrPages is the most pages read with OCR in one PDF. It keeps the cost of a long scanned document predictable.

Output example

{
  "url": "https://arxiv.org/pdf/1706.03762",
  "status": "ok",
  "pageCount": 15,
  "wordCount": 6462,
  "looksScanned": false,
  "ocrPageCount": 0,
  "ocrConfidence": null,
  "title": "Attention Is All You Need",
  "text": "Attention Is All You Need\n..."
}

Pricing

$5.00 per 1,000 PDFs ($0.005 each). One charge per PDF whose text was extracted, whatever its length.

OCR, when switched on, adds $8.00 per 1,000 scanned pages ($0.008 each). Pages with stored text are never charged for OCR.

200 ordinary PDFs cost $1. A 30-page scanned contract read with OCR costs $0.245: $0.005 for the PDF and $0.24 for its pages.

Each run also has a start charge of $0.00005. There is no separate charge for computing time. Apify's paid plans get 10% to 30% off these prices; the Pricing tab shows yours. Apify's free plan includes $5 of credit a month, which covers 1,000 ordinary PDFs or about 600 scanned pages.

Limits

• Without OCR it reads only the text stored in the PDF. Scanned documents then come back with empty text and looksScanned set to true.
• OCR reads printed text in the twelve languages listed above. It does not read handwriting.
• Files up to 40 MB. Text beyond 4 million characters is cut and the item is marked truncated.
• Tables come out as lines of text, not as structured rows and columns.

Questions

Can I run it on a schedule?

Yes. Save your input as a task and add an Apify schedule, for example every morning.

How do I get the results into Excel or Google Sheets?

Every run's results can be downloaded as Excel, CSV or JSON from the run's Output tab. Apify's integrations with Make, Zapier, n8n and Google Sheets can pass them on automatically.

Can I call it from my own code or from an AI agent?

Yes. The API tab has ready-made code for Python, JavaScript and cURL, and the tool can be connected to AI agents through Apify's MCP server.

Does OCR cost anything if my PDFs are not scanned?

No. With OCR switched on, only pages that have no stored text are read and charged.

Something looks wrong or is missing

Open a report under the Issues tab and include the input you used.