PDF to Text Extractor - Public PDF URLs, Page by Page

Convert PDF to text: extract searchable text from public PDF URLs, one dataset row per page. Bounded isolated parser, page limits, no OCR or login.

Quick start

Run the defaults or replace urls with public PDF document URLs you are authorized to process. The default is Mozilla PDF.js basicapi.pdf (Apache-2.0 test repository). No proxy, browser, account or API key is required.

{
  "urls": [
    "https://raw.githubusercontent.com/mozilla/pdf.js/master/test/pdfs/basicapi.pdf"
  ],
  "maxResults": 3,
  "maxPages": 3
}

Pricing

PPE result: $20 per 1,000 pages ($0.02/page). First 25 rows per run are free; only subsequent successful rows are charged. Errors are recorded in SUMMARY, never billed. Stops when the event spending limit is reached. No startup event or usage surcharge.

Inputs

• urls: 1-50 public HTTP(S) URLs. Check each source terms and reuse rights first; prohibited, login and private-network resources are not supported. Default: ['https://raw.githubusercontent.com/mozilla/pdf.js/master/test/pdfs/basicapi.pdf'].
• maxResults: Maximum successful dataset rows across all sources. Default: 100.
• keyword: Optional case-insensitive literal text filter; no regular expressions. Default: ``.
• delayMillis: Polite request delay, increased to honor robots crawl-delay. Default: 1000.
• maxResponseMb: Hard download size limit; oversized resources fail without bypassing. Default: 5.
• maxPages: Only the first N pages are parsed; totalPages retains source document count. Default: 20.
• maxCharsPerPage: Truncate long page text and flag textTruncated. Default: 20000.

Dataset

One row per parsed PDF page (including pages with no selectable text).

• sourceUrl (string): sourceUrl from the public source; absent source values may be null.
• pageNumber (integer): pageNumber from the public source; absent source values may be null.
• totalPages (integer): totalPages from the public source; absent source values may be null.
• text (string): text from the public source; absent source values may be null.
• textTruncated (boolean): textTruncated from the public source; absent source values may be null.
• hasText (boolean): hasText from the public source; absent source values may be null.
• fetchedAt (string): fetchedAt from the public source; absent source values may be null.

JSON/CSV/Excel export is provided by Apify datasets. SUMMARY gives total rows, source errors and stopReason. STATE supports migration resume; exactly-once delivery across a process failure between dataset write and state persistence is not guaranteed.

Limits and safety

Robots checked before every source/redirect; fail closed when unresolved. HTTP(S) public destinations only, DNS pinned to validated public addresses; private/reserved IPs blocked. 1-second minimum host delay, bounded requests and response sizes; no CAPTCHA, login or access-control bypass. No contacts/author-profile extraction. User is responsible for source terms, privacy and redistribution rights.

No OCR, image rendering, tables/layout reconstruction, attachments, JavaScript execution or password decryption. Scanned pages return hasText:false; textTruncated flags character caps. Each PDF is parsed in a memory-limited worker with a 15-second deadline. Page and character limits apply; up to 50 URLs. This is a text utility, not a document search index.