Enterprise-grade accuracy & speed

Extract text, forms, & tables at a fraction of the cost

A high-performance, cost-effective alternative to AWS Textract and Google Document AI. Send any image or PDF and get back clean, structured JSON in milliseconds.

Get your API key See pricing

How It Works

Simple, powerful, developer-friendly API

Sign up, generate an API key from your dashboard, and start sending POST requests. We handle layout parsing, confidence checking, model escalation, and parallel PDF processing behind a single simple endpoint.

Extract structured tables and form fields automatically
Split and analyze multi-page PDFs in parallel
Per-page billing with transparent pricing tier structures

# Analyze document (bytes as Base64)
POST /v1/document/analyze
Authorization: Bearer ocrg_live_key
Content-Type: application/json

{
  "document": { "bytes": "iVBORw0KGgoAAA..." },
  "featureTypes": ["FORMS", "TABLES"]
}

{
  "pages": [
    {
      "pageNumber": 1,
      "confidence": 0.994,
      "tables": [
        {
          "rows": [
            ["Description", "Qty", "Price"],
            ["API Processing", "1,000", "$1.50"]
          ]
        }
      ],
      "fields": {
        "InvoiceDate": "2026-07-13",
        "Total": "$1.50"
      }
    }
  ]
}

Text, forms & tables

Send any scanned image or PDF, get back plain text, key-value form fields, and table rows already parsed — no manual mapping required.

Multi-page PDFs

PDFs are automatically split page by page and processed in parallel, with per-page billing and confidence scoring.

Confidence-based review

Low-confidence pages are automatically escalated to a larger vision model, and flagged for human review if still uncertain.

Built for teams outgrowing Textract's pricing

AWS Textract charges up to $65 per 1,000 pages for combined form and table extraction. We charge a fraction of that, for the same structured output.

Compare prices & tier structures