Identity ~30s typical

Document Extraction

Extract text and key data from documents automatically with fast, accurate OCR.

Get API Request Demo

POST /v1/document/ocr

# Request
"input": "File1.pdf"

# Response fields
{
  "status_code": 200,
  "message": Success,
  "type": Classification,
  "document_type": RC,
  "fields": …,
  "confidence": …,
  "processed_at": …
}

~30s typical · sandbox available

Overview

What this endpoint does

Upload a document and get structured data back. The endpoint reads the file, identifies what kind of document it is, and extracts the key-value pairs configured for that type — no template to maintain per document layout.

Responses classify the document type — RC, PAN, Aadhaar, invoice and others — and return the configured keys alongside a per-field confidence score, so your workflow can route low-confidence extractions to a human instead of accepting them silently.

Response

What comes back

Every field is returned structured and timestamped, ready to store against the record it belongs to. Nothing requires parsing a photograph or interpreting free text.

See full documentation

Document OCR response fields
Field	Description
`status_code`	HTTP-style status for the extraction
`message`	Success, or the reason extraction failed
`type`	Whether the call classified, extracted, or both
`document_type`	Document class detected — RC, PAN, Aadhaar, invoice and so on
`fields`	Extracted key-value pairs for the configured keys
`confidence`	Per-field extraction confidence
`processed_at`	Server-side timestamp of the extraction

Use cases

Where teams use it

Onboarding document capture

Read the document a customer uploads instead of asking them to retype what it already says, and keep the typed result against the application.

Invoice and billing intake

Pull line items, totals and dates off supplier invoices so finance reconciles records rather than rekeying them.

Vehicle document processing

Extract registration details from an uploaded RC image, then pass them straight into an RC verification call.

Back-office digitisation

Convert a backlog of scanned paperwork into structured records, with confidence scores flagging the pages that need a human.

In context

Sectors that rely on this check

Banking & NBFC

Accelerate account opening and loan origination while maintaining rigorous compliance and audit controls.

Insurance

Generate quotes using verified vehicle data in real time, and settle claims based on validated records—not photographs alone.

Fintech

Complete customer onboarding in a single session while meeting the verification and compliance standards required by regulators and partners.

Also available

Related endpoints

Aadhaar–PAN Link Status

Confirm whether a PAN is linked to Aadhaar before a transaction depends on it — and fix the problem early if not.

Explore the API

PAN Verification

Validate a PAN and confirm the holder name matches your record, before it propagates into billing and tax.

Explore the API

Try Document OCR in the sandbox

Keys the same day, realistic payloads including failure cases, and engineers on the other end of the channel.

Get API Talk to sales