Image to Text OCR – 100+ Languages, Scanned PDFs, Word Boxes

Extract text from photos, scans, screenshots, receipts, invoices and scanned PDFs – in bulk, in 100+ languages, with a confidence score for every result.

• 🌍 100+ languages (Tesseract OCR) – English, German, French, Spanish, Polish, Russian, Arabic, Chinese, Japanese, Korean, Hindi… and mixes like eng+deu.
• 📑 Scanned PDFs and multi-page TIFFs – every page becomes one result.
• 🔄 Auto-rotation of pages scanned sideways or upside down.
• 📍 Optional line and word bounding boxes (x, y, width, height) – for highlighting, redaction or reading form fields.
• 🔗 Google Drive, Dropbox and GitHub share links work; you can also upload files.
• 💸 $0.002 per image or page. Failed files are free.

Use cases

• Read receipts and invoices before sending them to an LLM or accounting system.
• Make scanned contracts, letters and archives searchable.
• Pull text from screenshots, memes, slides and social media images for analysis or moderation.
• Extract product labels, signs and posters from photos.
• Give an AI agent eyes for text in images (works over MCP).

How to use

1. Paste image or PDF URLs (or upload files).
2. Set Languages (default eng).
3. Click Start – one dataset item per image / page.

Input example

{
  "urls": [
    "https://raw.githubusercontent.com/tesseract-ocr/test/main/testing/phototest.tif",
    "https://example.com/scanned-invoice.pdf"
  ],
  "languages": "eng",
  "includeLines": true
}

Output example

{
  "url": "https://raw.githubusercontent.com/tesseract-ocr/test/main/testing/phototest.tif",
  "fileName": "phototest.tif",
  "fileType": "image",
  "status": "ok",
  "text": "This is a lot of 12 point text to test the\nocr code and see if it works on all types\nof file format.\n\nThe quick brown dog jumped over the\nlazy fox. ...",
  "confidence": 95.5,
  "words": 60,
  "characters": 285,
  "width": 640,
  "height": 480,
  "rotation": 0,
  "languages": "eng",
  "lines": [
    {"text": "This is a lot of 12 point text to test the", "confidence": 95.5, "x": 36, "y": 92, "width": 544, "height": 30}
  ]
}

For PDFs each page is a separate item with page and pages. Coordinates are in pixels of the original image (PDF pages: rendered at 300 DPI; rotated images: after rotation).

Tips for best results

• Resolution matters: text should be at least ~20 px tall. Small images are upscaled automatically.
• Pick the right language – wrong language = garbled accents.
• Layout mode: Single column for receipts, Sparse for screenshots, forms and signs, Single line for cropped labels.
• Handwriting is not supported (printed text only).

Pricing

Event	Price
Image or PDF page read	$0.002

1,000 images cost $2. Apify platform usage is included. Set a maximum cost per run in the run options if you like.

Use it from code and AI agents

curl -X POST "https://api.apify.com/v2/acts/kris_fili~image-to-text-ocr/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"urls": ["https://example.com/receipt.jpg"], "languages": "eng", "pageSegmentation": "single-column"}'

from apify_client import ApifyClient
client = ApifyClient("YOUR_TOKEN")
run = client.actor("kris_fili/image-to-text-ocr").call(run_input={"urls": ["https://example.com/receipt.jpg"]})
print(next(client.dataset(run["defaultDatasetId"]).iterate_items())["text"])

MCP: https://mcp.apify.com?tools=kris_fili/image-to-text-ocr – lets Claude, Cursor and other AI agents read text from images.

Need Word, PowerPoint or Excel files as Markdown? Use Document to Markdown. Need screenshots of web pages? Use Website Screenshot & PDF.

FAQ

Is my data private? Images are processed only inside your run and results stay in your Apify storage.

What about tables? OCR returns text line by line. For PDFs that have a text layer, Document to Markdown extracts real tables.

Can it read CAPTCHAs? No, and it is not meant to.

Questions? Use the Issues tab.