PDF Text Extractor: Text, Pages & Metadata from PDF (OCR)

What does PDF Text Extractor do?

PDF Text Extractor gets the text out of any PDF and gives it back as clean JSON, CSV or Excel: the full text, the text of each page, the page count and the document's title, author, language and dates. It reads native and scanned PDFs (OCR for pages without a text layer) and also Word (DOCX), PowerPoint (PPTX), Excel, web pages and images.

Paste PDF URLs (Google Drive, Dropbox and OneDrive share links work) or upload files, and get one row per document. No API key, no browser, and you pay only per document processed, however many pages it has.

It runs on the Apify platform, so you also get an API, scheduling, integrations (Google Sheets, Make, Zapier, n8n) and access for AI agents through the Apify MCP server.

Why use it?

• 📄 Bulk PDF to text: contracts, reports, invoices, research papers, government forms, thousands at a time.
• 🔎 Text per page: search, cite or split documents by page.
• 🖨️ Scanned PDFs: pages without text are read with OCR (7 languages), and only those pages are billed as OCR.
• 🏷️ Metadata included: title, author, subject, creator, creation and modification dates, page count and detected language.
• 🧹 Clean text: repeated headers and footers removed, two-column layouts read in the right order.
• 🤖 For AI and RAG: switch on markdown or chunks to get LLM-ready Markdown or pieces with page numbers.

How to extract text from a PDF

1. Click Try for free.
2. Paste one or more PDF (or DOCX/PPTX/XLSX/HTML) URLs, or upload files.
3. Click Start.
4. Read the Documents view, or download the dataset as JSON, CSV or Excel.

Input

Field	Description	Default
sources	Document URLs (PDF, DOCX, PPTX, XLSX, HTML, images; share links work)	sources or base64Files
base64Files	Files sent inline (for APIs and AI agents)	-
outputs	text, pages, markdown, tables, chunks	text, pages
ocr	auto (only pages without text), always or never	auto
pageRange	e.g. 1-5, 8	all pages
pdfPassword	Password of encrypted PDFs	-
followDocumentLinks	Also extract the documents linked on a web page	false

{
  "sources": [{ "url": "https://www.irs.gov/pub/irs-pdf/fw9.pdf" }]
}

Output

One row per document. Real result for the IRS Form W-9 (October 2026; text and pages shortened):

{
  "source": "https://www.irs.gov/pub/irs-pdf/fw9.pdf",
  "status": "ok",
  "fileType": "pdf",
  "pageCount": 6,
  "title": "Form W-9 (Rev. March 2024)",
  "author": "SE:W:CAR:MP",
  "language": "en",
  "characters": 37694,
  "ocrPages": 0,
  "text": "W-9\nRequest for Taxpayer\nForm Give form to the\n(Rev. March 2024) Identification Number and Certification requester. …",
  "pages": [
    { "page": 1, "text": "W-9\nRequest for Taxpayer …", "ocr": false },
    { "page": 2, "text": "Form W-9 (Rev. 3-2024) Page 2\nmust obtain your correct taxpayer identification number (TIN) …", "ocr": false }
  ],
  "metadata": {
    "subject": "Request for Taxpayer Identification Number and Certification",
    "creator": "Designer 6.5",
    "created": "2024-03-06T08:18:13",
    "modified": "2024-03-06T08:18:13",
    "pageCount": 6
  }
}

You can download the dataset in various formats such as JSON, HTML, CSV, or Excel.

Data fields

Field	Description
text	The whole document as plain text
pages[].page / pages[].text / pages[].ocr	Text of each page, and whether it was read with OCR
pageCount, characters	Size of the document
title, author, language	From the document metadata (language is detected from the text)
metadata	Subject, creator, producer, creation and modification dates
ocrPages	Pages read with OCR (billed as OCR pages)
status, warnings	ok or error, and anything worth knowing (for example, no text found)

How much does it cost to extract text from PDFs?

Event	Price
Run start (per GB of memory)	$0.001
Document processed (any number of pages)	$0.003
OCR page (only scanned pages without text)	$0.01

1,000 native PDFs cost about $3, whatever their length. Failed documents are free. Set Max cost per run and the Actor stops cleanly at that limit.

Use it with AI agents (MCP)

Add https://mcp.apify.com?tools=fguiraud/pdf-text-extractor to Claude, Cursor or any MCP client and ask: "Read this contract and list the termination clauses: https://example.com/contract.pdf".

Smallest useful input for an agent: {"sources": [{"url": "https://www.irs.gov/pub/irs-pdf/fw9.pdf"}]}.

Agent-friendly input: common field names such as url, urls, startUrls, query or keywords are accepted, and an empty input runs a small sample from the input form (with a note in the results) instead of failing.

Related tools

• PDF to Markdown Extractor & Document Parser: LLM-ready Markdown, RAG chunks and AI field extraction (invoices, forms), same engine.
• PDF Table Extractor: every table to Excel, CSV and JSON.

FAQ and limitations

• Scanned PDFs: OCR is automatic for pages without text; ocr: "never" turns it off, "always" forces it.
• Encrypted PDFs: enter the password in PDF password.
• Very large documents: results over 9 MB drop the per-page text first and keep the full text, with a warning.
• Found a problem or need a feature? Open an issue on the Issues tab. Replies within 48 hours.