PDF to Text Extractor (Bulk, with Metadata)

Turn a list of PDF links into clean text. You paste direct PDF URLs, and you get one dataset row per PDF with the full text, optional page-by-page text, and metadata: title, author, subject, creator, producer, creation and modification dates, page count, word count, character count and file size. It is built for feeding LLMs and RAG pipelines, search indexes, research and document analysis, without running your own PDF parsing code.

Extract text from PDF URLs in bulk

Give the actor a list of pdfUrls. It downloads each file, reads the text layer and returns a row per document. Several PDFs are processed in parallel (maxConcurrency, 1 to 10, default 5). Files that are too large, missing or not actually PDFs come back as rows with an error field and are not charged.

Input example

{
  "pdfUrls": [
    "https://arxiv.org/pdf/1706.03762",
    "https://example.com/reports/annual-2025.pdf"
  ],
  "perPage": true,
  "maxPages": 200,
  "maxFileSizeMb": 50,
  "maxConcurrency": 5
}

Field	Default	What it does
pdfUrls (required)		Direct links to PDF files
perPage	false	Adds a pages array with the text of each page
maxPages	500	Only extract the first N pages of each PDF
maxFileSizeMb	50 (max 200)	Skip PDFs larger than this
maxConcurrency	5 (1 to 10)	PDFs processed in parallel

PDF to text with metadata: output

One realistic result row (text shortened):

{
  "url": "https://arxiv.org/pdf/1706.03762",
  "pageCount": 15,
  "pagesExtracted": 15,
  "title": null,
  "author": null,
  "wordCount": 6120,
  "likelyScanned": false,
  "text": "Provided proper attribution is provided, Google hereby grants permission ...",
  "pages": [{ "page": 1, "text": "..." }]
}

Other fields in the dataset schema: subject, creator, producer, createdAt, modifiedAt (ISO timestamps), fileSizeBytes, charCount, contentType (only when the response was not a PDF) and error. Metadata fields are null when the PDF does not contain them, as with title and author above. pages appears only when perPage is on.

Use cases

• Build a RAG or LLM knowledge base from reports, papers and manuals. Turn on perPage to get page-numbered chunks you can cite.
• Bulk-index contracts, invoices and filings for full-text search.
• Extract PDF metadata (author, dates, page count) across a document library.
• Triage a PDF batch. likelyScanned flags files that are mostly images, so you know which need OCR before you send them to a language model.
• Word and page statistics for research corpora using wordCount, charCount and pageCount.

Pricing: how much does it cost to convert PDFs to text?

$0.003 per successfully extracted PDF, flat, regardless of page count (up to maxPages). No subscription.

• 100 PDFs = $0.30
• 1,000 PDFs = $3.00
• 10,000 PDFs = $30.00

Failed rows (HTTP errors, not a PDF, too large) are not charged. Apify platform compute usage is separate from the per-PDF price; check your run's cost in the console.

Limitations

• No OCR. Scanned or image-only PDFs have no text layer. They are flagged likelyScanned: true (fewer than about 5 words per page on average) and the text will be empty or very short.
• Password-protected PDFs are not supported and return an error row.
• Direct URLs only. The link must return the PDF file itself, not a web page that embeds a viewer. Each download times out after 60 seconds.
• Layout is not preserved. Text comes out in reading order; tables become plain text lines and figures are ignored.
• Size and page caps. Files above maxFileSizeMb are skipped, and only the first maxPages pages are read.
• Only process documents you are allowed to access and use.

FAQ

How do I extract text from a PDF URL? Put the direct link in pdfUrls and run the actor. The text field in the dataset holds the full document text.

Can I convert many PDFs to text at once? Yes. Add as many URLs as you need; they run in parallel up to maxConcurrency.

Does it OCR scanned PDFs? No. Scanned files are flagged likelyScanned: true so you can route them to an OCR tool.

Am I charged for failed downloads or bad files? No. Only successfully extracted PDFs are billed at $0.003 each.

How do I get text per page for citations? Set perPage to true; each row then includes pages with page numbers and text.

Can it read metadata like author and creation date? Yes: title, author, subject, creator, producer, createdAt and modifiedAt, when the PDF includes them.

Built and maintained with AI assistance. Problems or requests: use the Issues tab.

Related tools

Other actors by the same developer (same flat pay-per-result pricing, no subscription):

• Apple App Store Reviews Scraper (Multi-Country)
• Article Extractor – Clean Text & Markdown for LLM/RAG
• Company Jobs Scraper: Workday, Greenhouse, Lever, Ashby
• Bulk Domain Checker — WHOIS/RDAP, DNS, SPF/DMARC, SSL Expiry
• Bulk PageSpeed Insights & Core Web Vitals Checker
• Website SEO Audit Crawler
• Sitemap Extractor & Bulk URL Status Checker
• Website Tech Stack Detector (CMS, Framework, Analytics)
• Website Screenshot – Bulk Full Page PNG, JPEG & PDF

Integrations

Run it from the Apify API, a schedule, or no-code tools: the Apify apps for Zapier, Make and n8n can start any public actor ("Run Actor") and read its dataset. AI agents can call it through the Apify MCP server.