Docs
Get started

Getting Started

Introduction

LiveXFace is a REST API that adds face recognition capabilities to any application, regardless of the language or framework you use. Send an image, get back a result. No ML infrastructure to manage, no model training required.

What you can build

Access Control
Replace keycards or PINs with face-based door/gate access. Enroll employee faces once, identify in real time.
Identity Verification
Verify that a selfie matches a stored profile photo. Useful for KYC, onboarding, or re-authentication flows.
Attendance Systems
Automate check-in by recognizing faces from a camera feed without any manual interaction.
Watchlist Screening
Search an uploaded face against a curated list and receive a match score instantly.
Photo Deduplication
Find duplicate or near-duplicate face photos across a large media library.
Anti-Fraud
Block spoofing attempts (printed photos, screen replays) during account registration or login.

How it works

The API is organized around Face Collections: isolated groups of enrolled faces, one per application, tenant, or use-case. Here's the basic flow:

  1. Create a collection: a named bucket that holds all enrolled faces for a specific use-case.
  2. Enroll faces: upload a photo with your own identifier (e.g. a user ID). The API handles detection, quality checks, and storage automatically.
  3. Verify or identify: call /verify to check if two faces match (1:1), or /identify to find who a face belongs to from your enrolled set (1:N).

API capabilities

  • Face enrollment: Register people into a collection with custom metadata. Supports batch uploads.
  • 1:1 Verification: Compare two face images or verify a probe against an enrolled face ID.
  • 1:N Identification: Search your collection and return the top-K most similar faces with similarity scores.
  • Liveness detection: Passively detect spoofing attempts (printed photos, screen replays) before enrollment or as a standalone check.
  • Scoped API keys: Each integration gets its own key, scoped to a specific collection, with configurable rate limits and IP whitelisting.
  • Webhooks: Receive real-time HTTP notifications when faces are enrolled, identified, or liveness checks fail.

Try it before you write anything

The console has a playground covering every recognition endpoint. Point a camera at yourself or upload a file, and you will see the same response your code is going to get — useful for settling on a threshold before it is hard-coded anywhere.

The Playground page in the LiveXFace console. Tabs across the top read Compare, Verify, Identify, Liveness, Active Liveness, Live Recognition, Attributes and Cross-Search. The Compare tab is open, showing two capture panes, each offering Open Camera or a file upload.
One tab per endpoint. Each takes a live camera frame or a file, and prints the response underneath.

Authentication

One API key per integration, sent as an X-API-Key header. Issue keys from the console; your application never logs in. See Authentication for scopes, IP whitelisting and rate limits.

Base URL

https://api.livexface.com/api/v1

Self-hosted deployments swap the host for their own; every path below it stays the same.

Recognition endpoints are scoped to a collection:

POST /api/v1/collections/{collection_id}/identify
POST /api/v1/collections/{collection_id}/verify
POST /api/v1/collections/{collection_id}/faces

Works with any stack

LiveXFace is a plain HTTP REST API. It works from any language or framework that can make HTTP requests and handle multipart/form-data file uploads:

Node.jsPythonPHPGoRubyJavaSwiftKotlincURLPostman

See the Quick Start guide for code examples in multiple languages.