curl --request POST \
  --url https://extraction-api.nanonets.com/api/v2/classify/sync \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "input": "file://a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "classification_config": {
    "categories": [
      {
        "name": "Invoice",
        "description": "Bills and invoices"
      },
      {
        "name": "Receipt",
        "description": "Payment receipts"
      },
      {
        "name": "Contract",
        "description": "Legal agreements"
      }
    ],
    "mode": "document"
  }
}
'

{
  "success": true,
  "message": "Document classified successfully",
  "status": "completed",
  "result": {
    "filename": "invoice.pdf",
    "total_pages": 3,
    "mode": "document",
    "document_category": "Invoice",
    "confidence": 92.5,
    "reasoning": "Page 1: Contains invoice header with vendor details; Page 2: Itemized list of products with pricing"
  },
  "processing_time": 1.5,
  "filename": "invoice.pdf"
}

Document Classification

Synchronous Classification

Classify a document into predefined categories synchronously.

Modes:

document — Returns a single best-fit category for the whole document with confidence and reasoning.
split — Classifies each page individually and groups pages by category. A page can appear in multiple categories (multi-label).

Provide categories inline via classification_config or reference a saved classify config via config_id.

Tip: You can save and reuse classification settings (categories & mode) by creating a config via Create Config. Pass the returned config_id instead of repeating classification_config each time.

POST

api

classify

sync

curl --request POST \
  --url https://extraction-api.nanonets.com/api/v2/classify/sync \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "input": "file://a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "classification_config": {
    "categories": [
      {
        "name": "Invoice",
        "description": "Bills and invoices"
      },
      {
        "name": "Receipt",
        "description": "Payment receipts"
      },
      {
        "name": "Contract",
        "description": "Legal agreements"
      }
    ],
    "mode": "document"
  }
}
'

{
  "success": true,
  "message": "Document classified successfully",
  "status": "completed",
  "result": {
    "filename": "invoice.pdf",
    "total_pages": 3,
    "mode": "document",
    "document_category": "Invoice",
    "confidence": 92.5,
    "reasoning": "Page 1: Contains invoice header with vendor details; Page 2: Itemized list of products with pricing"
  },
  "processing_time": 1.5,
  "filename": "invoice.pdf"
}

Authorizations

Authorization

string

header

required

API key as Bearer token: Authorization: Bearer YOUR_API_KEY

Body

application/json

input

required

file://<uuid> URI or public URL. Async accepts a list (max 50).

config_id

string

config://<uuid> of a saved classify config

classification_config

object

Show child attributes

Response

Classification completed

success

boolean

required

message

string

required

record_id

string

file_id

string

config_id

string

status

enum<string>

Available options:

completed,

processing,

failed

result

object

Classification result (shape depends on mode)

Option 1
Option 2

Show child attributes

processing_time

number

Time in seconds

filename

string

file_size

integer

Size in bytes

Asynchronous Extraction Batch Classification

⌘I

File Upload

Document Parsing

Document Extraction

Document Classification

Configuration Management

Synchronous Classification

Authorizations

Body

Response