PDF Text Extractor — Extract Text from Any PDF File

Extract text from PDF files with OCR support for scanned documents and image-based PDFs. Supports direct URLs and cloud storage links.

🎯 What It Does

• Extract text from any PDF — digitally created or scanned
• Cloud storage support — Google Drive, Dropbox, and OneDrive share links
• OCR fallback — automatically runs Tesseract OCR on pages with no embedded text
• Multi-language OCR — 100+ languages supported
• Mistral AI OCR fallback — when both pdf.js-extract and Tesseract fail, optionally use Mistral OCR for state-of-the-art document understanding (tables, equations, complex layouts)
• Page limiting — optionally cap extraction to a specific number of pages
• Structured output — extracted text plus metadata (page count, source type, file size, timestamp)

📥 Input

{
  "pdfUrl": "https://drive.google.com/file/d/FILE_ID/view?usp=sharing",
  "maxPages": 0,
  "ocrFallback": true,
  "ocrLanguage": "eng",
  "mistralApiKey": "your-mistral-api-key"
}

Field	Type	Default	Description
pdfUrl	string	—	PDF URL or cloud storage share link (Google Drive, Dropbox, OneDrive)
maxPages	integer	0	Max pages to extract. 0 = all pages
ocrFallback	boolean	true	Run Tesseract OCR on pages with <50 chars of embedded text
ocrLanguage	string	eng	Tesseract language code (e.g. fra, deu, chi_sim)
mistralApiKey	string	—	Optional Mistral AI API key. When provided, if both pdf.js-extract and Tesseract fail to produce meaningful text, the PDF is sent to Mistral OCR for premium document understanding

📤 Output

{
  "originalPdfUrl": "https://drive.google.com/file/d/FILE_ID/view",
  "processedPdfUrl": "https://drive.google.com/uc?export=download&id=FILE_ID",
  "extractedText": "Full text content extracted from the PDF...",
  "pageCount": 12,
  "extractedPages": 12,
  "fileSizeBytes": 1048576,
  "sourceType": "google-drive",
  "ocrApplied": true,
  "mistralFallbackApplied": false,
  "timestamp": "2026-06-05T07:00:00.000Z",
  "success": true
}

🔍 How It Works

1. Downloads the PDF via HTTP with retry logic (3 attempts, exponential backoff)
2. Extracts embedded text using pdf.js-extract (fast for digital PDFs) — Stage 1
3. Tesseract OCR fallback (when enabled): pages with <50 chars are rendered to PNG at 300 DPI via pdftoppm, then processed with Tesseract OCR — Stage 2
4. Mistral OCR premium fallback (when mistralApiKey is set): if Stages 1 and 2 produce fewer than 200 characters total, the PDF URL is sent to Mistral AI's OCR API for state-of-the-art document understanding — Stage 3
5. Returns structured JSON with extracted text and metadata

💰 Pricing

Pay-per-event — charged only on successful extractions.

Event	Price	Trigger
pdf-processed	$0.005	Per successfully processed PDF
page-extracted	$0.0005	Per page (only when extractedPages > 1)

🚀 Use Cases

• Document processing — invoices, contracts, reports, scanned paper copies
• Research — academic papers, white papers, archival PDFs
• Data pipelines — feed PDF content into NLP or search systems
• Content management — index PDF archives for full-text search
• Automation — process PDFs at scale via Apify API, Zapier, or Make

⚡ Tips

• Scanned PDFs: ocrFallback: true is enabled by default — works out of the box
• Large PDFs: set maxPages to limit processing time and cost
• Non-English docs: set ocrLanguage to the matching Tesseract language code
• Failed extractions: not charged — error details returned in errorMessage field