PDF Extractor - Text, Tables, Metadata & Links

Turn a list of PDF links into clean, structured data. For each PDF you get the full text, the text per page, document metadata (title, author, dates, producer), every hyperlink, the outline (bookmarks), and - optionally - tables reconstructed into rows and columns.

Pure pdfjs: no browser, no third-party OCR service, no anti-bot target. Runs in seconds and costs a fraction of a cent per document.

What you get per PDF

Field	Description
text	Full document text, with line and column layout preserved (tables stay readable, not word-soup)
pages[]	Each page's text separately (optional)
pageCount wordCount charCount sizeBytes	Size and length metrics
title author subject keywords creator producer pdfVersion isEncrypted creationDate modificationDate	Document metadata
links[] linkCount	Every http(s) link - from link annotations and from the text
outline[]	Bookmarks / table-of-contents entries, when present
tables[]	Best-effort row/column grids per page (opt-in via Extract tables)
status error	ok, or failed with the reason (bad link, not a PDF, too large)

Input

{
  "pdfUrls": [
    "https://arxiv.org/pdf/1706.03762",
    "https://www.w3.org/WAI/ER/tests/xhtml/testfiles/resources/pdf/dummy.pdf"
  ],
  "extractTables": true,
  "includePageText": true
}

• PDF URLs - direct links to .pdf files, one per line. Leave empty for a small free sample.
• Import PDF URLs from a file - paste a whole list, or link a public .txt/.csv, a Google Sheet/Drive share link, or an Apify key-value-store record. Every common messy-file quirk (blank lines, # comments, quotes, extra CSV columns) parses.
• Output fields - tick only the fields you want for a lean export.
• Max pages / Max PDF size - cost controls for very large documents.

Notes on accuracy

• Text extraction reads the PDF's own text layer, so it is exact for digitally-created PDFs. Scanned image-only PDFs have no text layer and will return little or no text (this actor does not run OCR).
• Table detection is best-effort: PDFs carry no standard table markup, so tables are reconstructed from text position. Ruled and well-aligned tables come out cleanly; free-form layouts may not.

Use with AI assistants (MCP)

This actor is available through the Apify MCP server, so an AI assistant or agent can call it as a tool - hand it PDF links and get back structured text, metadata and tables it can reason over (summarise a report, pull figures from a filing, index a document set).

Agent-ready: autonomous payments (x402 & Skyfire)

This actor is agent-ready - AI agents can discover it, run it, and pay for it autonomously, with no Apify account and no human in the loop. It uses pay-per-event pricing and limited permissions, so it qualifies for Apify's agentic-payment standards:

• x402 - an open, HTTP-native payment protocol. Agents pay per run in USDC on the Base network directly through the Apify MCP server - no account, no API key.
• Skyfire - agent-to-service payments for fully autonomous AI-agent workflows.

Building an AI agent, MCP tool, or autonomous data pipeline? This scraper is ready to plug in and pay as it goes.