OCR API
OCR API for PDFs and images
Extract text, blocks, coordinates and confidence from scans and native files.
INPUT scan.pdf OUTPUT text · blocks · confidence
{
"text": "Invoice #1842",
"blocks": [{ "page": 1 }],
"confidence": 0.98
}WHAT YOU GET
Text, blocks and confidence
Clean text plus page-aware structure
PDF, PNG, JPEG and WEBP
One upload contract
Bounding boxes
Use when coordinates matter
Language and page results
Inspect document-level detail
ONE PREDICTABLE REQUEST
Send a document. Get the same response format every time.
curl -X POST https://api.startocr.com/v1/ocr \ -H "Authorization: Bearer $STARTOCR_API_KEY" \ -F file=@scan.pdf
BUILT FOR PRODUCTION
Works beyond the demo.
Async jobs, page-level results, retries, webhooks, request logs and usage tracking.
Same response every time
Use the same field names across the API
Smart processing
Balances speed and accuracy
Easy controls
Check errors and costs in one place