BlinkID Verify API v2 Reference

Open API v2 Reference

Simpler API

The v2 API is simpler (but at the same time comes with more information) and more intuitive than the v1 API.

We've designed it to be less nested, easier to integrate, and easier to reason about. One guiding principle of the v2 API was grouping, which enables you to quickly get answers to only the questions you care about (at a given time).

The overall fraud verdict is one top level object, images are another, detailed fraud check information is yet another, etc. The result contains the following groups:

• processingStatus - single piece of meta information on the entire verification process: did it complete, and if it did not fully complete, what happened?
• verification - information about the actual verification: was the document genuine or not?
• checks - detailed breakdown of all fraud checks: if the document got rejected, why?
• processIndicators - meta information about potential capture+extraction+verification process: were the conditions suitable for verification or not?
• images - the images that were used for verification
• extraction - all the personal data that was extracted from the document
• messages - human readable messages (and codes) meant for integrators: if the verification performed unexpectedly, why and how to mitigate it?
• optionsUsed - the options that were used for this particular verification: which checks were enabled, which checks were disabled, etc.
• useCaseUsed - what was the UseCase that was defined for this particular verification?
• runtime - meta information about the process: how long did it take, what versions were used, traceId, etc.

Additional Features

The v2 API is our current long term support API, and as such is where we add new features that require API additions.

• Recommended outcome - a new field that advises you what to with a specific verification result, depending on your use case. For example, the fraud verdict might be FAIL, but we also deem the image quality low enough to bring that into question. Instead of having to implement this kind of logic, you can use the RecommendedOutcome

• Use case - a high level object that allows you to define your use case with a few simple parameters from which we automatically determine optimal options. You're also able to override individual options.

v1 vs v2 Comparison

Feature	v1	v2
Breakdown per check	✅	✅
Fraud verdict	✅	✅
Extracted data	✅	✅
Multipart	✅	✅
UseCase configuration	❌	✅
RecommendedOutcome	❌	✅
Iterable checks	❌	✅

Schema Files

If you're just looking for raw schema files, you can download them here:

• Request Schema
• Response Schema