Skip to main content
Requirements:
  • API key pair
  • Direct Processing feature enabled on your contract (contact iDenfy to activate)
Limitations:
  • Does not support 3D liveness detection or short photo sequences. Use a different integration type if these are required.
  • Maximum request size: 20 MB (total across all images).
  • Cannot be used to update an existing verification. Use the Request Update feature or re-verify using the same clientId via session creation.
  • Digital ID types are not supported and will be rejected.

How It Works

Direct processing lets you submit document images in a single POST request. The API responds immediately — the actual OCR, face matching, fraud checks, and AML screening all run asynchronously after the response. Results are delivered via webhook.

Endpoint


Request Parameters

Required

The verification session token generated beforehand via session creation.
  • Must be valid, unused, and belong to the authorized partner.
  • Deactivated immediately after the request — cannot be reused.
  • Must not have questionnaire answers (kyc_answers) attached.
  • NFC must not be required for the session.
ISO 3166-1 alpha-2 country code (e.g. LT, US, DE). Converted to uppercase automatically. Must be a valid recognized country code.
A dictionary of Base64-encoded images, keyed by side:Supported formats: PNG, JPG, JPEG, PDF. Encrypted or password-protected PDFs are rejected.

Optional

Specifies the document being submitted. Valid values: ID_CARD, PASSPORT, DRIVER_LICENSE, RESIDENCE_PERMIT.If omitted, the system automatically detects the document type from the FRONT image. If auto-detection fails, the request is rejected with an error.
When documentType is provided, BACK becomes required if that document type has a back side.
Set to true to skip photo quality checks (blur, glare, document detection). Useful when a valid photo fails standard detection.Does not skip data extraction or fraud checks — only the photo validation step.

Verification Types

The session tokenType set during session creation determines what images are expected:
  • IDENTIFICATION (Default): Requires a FACE image. Face matching runs against the document photo.
  • DOCUMENT: Document-only. Face step is removed automatically — no FACE image or liveness check required. Result is marked FACE_NOT_CHECKED.

Additional Steps

Some verification flows require extra document checks beyond the main identity document — for example, a utility bill for address verification. These are configured as additional steps on the token at session creation, not in the /v2/process call itself. How it works:
  1. Token creation — the partner sets up the token with UTILITY_BILL as a required additional step, and optionally passes a reference address to compare against via additionalData:
  2. Direct processing call — the partner includes the utility bill image inside the images object alongside the main document images. No extra fields are needed in the /v2/process request body.
  3. Processing — depending on the step configuration, the system will:

Conditions and Rules


Request Examples

With explicit document type:
With auto document type detection (documentType omitted):

Response

A successful request returns HTTP 200 with no response body. Processing continues asynchronously.

Error Response

Failed requests return a JSON body identifying the problem:
A 200 OK with an error message body may indicate the document was not detected in the photo. If this happens:
  • Add "skipAnalysis": true to the request, or
  • Use a clearer photo with a fully visible document.
See the full list of error messages.

What Happens After

Once the request is accepted, processing runs asynchronously:
  1. OCR extracts document fields — name, date of birth, expiry, document number, nationality, and more.
  2. Face matching runs if a FACE image was provided.
  3. Fraud checks and AML/sanctions screening run automatically.
  4. Results are delivered via webhook to your configured endpoint.
See Webhooks for payload details.