API Docs — OCR Engine

Authentication

All requests are authenticated with a bearer token. Pass your API key in the Authorization header. Request a key from the contact form.

header

Authorization: Bearer YOUR_API_KEY

Extract endpoint

Send a PDF as multipart form data along with a fields object describing what you want back.

POST https://api.ocrengine.io/v1/extract

curl https://api.ocrengine.io/v1/extract \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@invoice.pdf" \
  -F 'fields={
      "invoice_number": "string",
      "total_amount":   "number",
      "due_date":       "date",
      "line_items":     "array"
    }'

Defining parameters

The fields object is where you describe the data you want. Each key is the name you'll get back in the response; each value is the expected type. The engine finds and returns only those fields.

Parameter	Type	Required	Description
`file`	file (PDF)	required	The PDF to read. Native or scanned, up to 50 MB.
`fields`	object	required	Map of field name → expected type.
`language`	string	optional	Hint such as `en` or `auto` (default).
`confidence`	boolean	optional	Include a confidence score per field. Default `true`.
`webhook_url`	string	optional	Process asynchronously and POST the result here.

Field types

Use these values inside the fields object to describe each parameter.

Type	Returns	Example
`string`	Text value	`"INV-0425"`
`number`	Numeric value	`1840.00`
`date`	ISO-8601 date	`"2026-07-15"`
`boolean`	true / false	`true`
`array`	List of values or rows	`[ … ]`
`object`	Nested key/value group	`{ … }`

Response

You get back a JSON object containing exactly the parameters you defined. When confidence is enabled, each field is paired with a score between 0 and 1.

200 OK response.json

{
  "status": "completed",
  "data": {
    "invoice_number": "INV-0425",
    "total_amount": 1840.00,
    "due_date": "2026-07-15",
    "line_items": [
      { "description": "Design retainer", "amount": 1500.00 },
      { "description": "Hosting", "amount": 340.00 }
    ]
  },
  "confidence": {
    "invoice_number": 0.99,
    "total_amount": 0.97,
    "due_date": 0.95
  }
}

Errors

The API uses conventional HTTP status codes. Error responses include a machine-readable code and a human-readable message.

Status	Code	Meaning
`400`	`invalid_request`	Malformed `fields` object or missing file.
`401`	`unauthorized`	Missing or invalid API key.
`413`	`file_too_large`	PDF exceeds the 50 MB limit.
`422`	`extraction_failed`	Document could not be read.
`429`	`rate_limited`	Too many requests — slow down or contact us.

Need something not covered here? Reach out and we'll help.