Mistral Document AI

Enterprise-grade OCR powered by Mistral's document understanding models.

Capabilities

• Text Extraction: Accurate text from scanned documents
• Table Detection: Structured table extraction
• Handwriting: Handwritten text recognition
• Layout Analysis: Document structure understanding
• Multi-language: 100+ languages supported

Basic Usage

curl -X POST https://api.gateflow.ai/v1/data/ocr \
  -H "Authorization: Bearer gw_prod_..." \
  -F "file=@scanned_document.pdf"

Response:

{
  "document_id": "doc_123",
  "pages": [
    {
      "page_number": 1,
      "text": "Contract Agreement...",
      "tables": [...],
      "confidence": 0.95
    }
  ]
}

Configuration

{
  "ocr": {
    "provider": "mistral",
    "extract_tables": true,
    "detect_handwriting": true,
    "language_hint": "en",
    "output_format": "markdown"
  }
}

Table Extraction

{
  "tables": [
    {
      "id": "table_1",
      "rows": [
        ["Item", "Quantity", "Price"],
        ["Widget A", "100", "$10.00"],
        ["Widget B", "50", "$20.00"]
      ],
      "confidence": 0.92
    }
  ]
}

Pricing

Feature	Price
Standard OCR	$0.01/page
Table extraction	$0.02/page
Handwriting	$0.03/page

Next Steps

• Supported Formats - File types
• Table Extraction - Advanced tables