Process Document
The process endpoint is the core of Cardinal OCR. Upload documents and receive extracted text, structured data, images, and metadata in your preferred format.
Process Endpoint
https://api.trycardinal.ai/v1/processUpload a document file and receive processed OCR results with text extraction, structured data, and optional image/metadata extraction.
multipart/form-data when uploading files to the process endpoint.Parameters
fileThe document file to process. Supports PDF, images, and other document formats.
output_formatOutput format for the processed document. Default: json
extract_imagesExtract images, charts, and visual elements from the document. Default: false
include_bounding_boxesInclude precise coordinate data for all text elements. Default: false
extract_metadataExtract barcodes, QR codes, and other metadata. Default: false
extract_signatureExtract detected signatures from the document. Default: false
Supported File Formats
Supported: PDF (.pdf)
Coming Soon: Excel sheets (.xlsx, .xls) support is currently in development.
Image Formats: JPEG (.jpg, .jpeg), PNG (.png), TIFF (.tif, .tiff)
Response Format
All successful responses return a JSON object with the processed document data, regardless of the chosen output format.
{
"status": "success",
"document_id": "doc_abc123",
"content": {
"text": "Extracted text content...",
"structured_data": {
// Structured data based on document type
}
},
"metadata": {
"page_count": 3,
"processing_time_ms": 2500,
},
"images": [
// Array of extracted images (if extract_images=true)
],
"bounding_boxes": [
// Array of coordinate data (if include_bounding_boxes=true)
],
"barcodes": [
// Array of detected barcodes/QR codes (if extract_metadata=true)
],
"signatures": [
// Array of detected signatures (if extract_signature=true)
]
}{
"error": {
"code": "invalid_file_format",
"message": "Unsupported file format",
"details": {
"supported_formats": ["pdf", "png", "jpg"]
}
}
}- • Simple documents: 1-30 seconds (depending on the document size)
- • Complex documents: 1-2 minutes (depending on the document size)
- • Large files (50MB): Couple of minutes
Examples
curl -X POST https://api.trycardinal.ai/v1/process \ -H "Authorization: Bearer YOUR_API_KEY" \ -F "file=@invoice.pdf" \ -F "output_format=json"
extract_images=true and extract_metadata=true and extract_signature=true for comprehensive data extraction.