Document OCR API — IteraTools

Why AI OCR?

Traditional OCR (like Tesseract) struggles with tables, multi-column layouts, handwriting, and low-quality scans. Mistral AI OCR understands document structure, preserves table formatting in Markdown, and handles PDFs with mixed text and images — including scanned Brazilian notas fiscais, contracts, and forms.

Quick Start — Image URL

curl -X POST https://api.iteratools.com/document/ocr \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"image": "https://example.com/invoice.jpg"}'

Quick Start — PDF URL

curl -X POST https://api.iteratools.com/document/ocr \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"pdf": "https://example.com/contract.pdf", "pages": [0, 1, 2]}'

Quick Start — Base64 Image

curl -X POST https://api.iteratools.com/document/ocr \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"image\": \"data:image/jpeg;base64,$(base64 -w0 invoice.jpg)\"}"

Response

{
  "ok": true,
  "data": {
    "text": "Invoice #1234\nDate: 2024-01-15\nTotal: R$ 450,00",
    "markdown": "# Invoice #1234\n\n| Item | Qty | Price |\n|------|-----|-------|\n| Widget | 2 | R$225 |\n\n**Total: R$ 450,00**",
    "tables": [
      {
        "headers": ["Item", "Qty", "Price"],
        "rows": [["Widget", "2", "R$225"]]
      }
    ],
    "pages": 1,
    "model": "mistral-ocr-latest"
  }
}

Request Body

Field	Type	Description
image	string	JPEG or PNG as base64 (data:image/... prefix) or public URL. Max 10MB.
pdf	string	PDF as base64 (data:application/pdf;base64,...) or public URL. Max 10MB.
pages	number[]	Optional. Array of 0-based page indices to process. Default: all pages.

Use Cases

📋 Extract data from scanned invoices and receipts (notas fiscais)
📊 Parse tables from PDFs into structured data
📝 Digitize handwritten or printed forms
📑 Process multi-page contracts and legal documents
🏦 Read bank statements and financial reports
🔍 Make scanned PDFs searchable

vs /image/ocr

Feature	/image/ocr (Tesseract)	/document/ocr (Mistral AI)
Tables	❌ Plain text	✅ Structured Markdown
PDF support	❌ Images only	✅ Native PDF
Multi-column layouts	❌ Often garbled	✅ Preserves structure
Price	$0.002/req	$0.015/req
Best for	Simple text images	Complex docs, PDFs, forms

Pricing

$0.015 per request via x402 micropayment on Base (USDC). No monthly fees. Pay per use.