Voice Authentication SDK
Enterprise-grade voice biometric SDK with support for 8+ programming languages. Implement secure voice authentication in minutes, not months.
Built for Modern Development
Sub-second voice verification with optimized algorithms
End-to-end encryption with AES-256-GCM biometric protection
Support for 50+ languages with accent adaptation
Process voice data locally for enhanced privacy
Seamlessly scale across regions with our cloud infrastructure
GDPR, CCPA, and SOC 2 compliant out of the box
Up and running in minutes
Install the SDK, enrol a voice, authenticate — three steps to production-ready voice biometrics.
Install
npm install @voiceidvault/sdkEnrol a voice, then authenticate
import { VoiceIDVaultClient } from '@voiceidvault/sdk';
const client = new VoiceIDVaultClient({
apiKey: 'viv_your_api_key',
});
// Step 1: Initialise an enrollment session
const session = await client.initEnrollment('user_123', '+12125551234');
// Step 2: Enrol the user's voice (Blob from MediaRecorder, 16 kHz mono WAV)
const enrolment = await client.enrollVoice(
session.sessionId,
'user_123',
wavBlob, // Blob | ArrayBuffer
{ phrase: 'My voice is my password', sampleRate: 16000 }
);
console.log(enrolment.voiceprintId); // 'voice_a1b2c3...'
console.log(enrolment.qualityScore); // 0.92
// Step 3: Complete enrollment
await client.completeEnrollment(session.sessionId);
// Later — authenticate by voice
const result = await client.authenticateVoice(wavBlob, { userId: 'user_123' });
if (result.verified) {
console.log('Confidence:', result.confidence); // 0.97
console.log('Trust score:', result.trustScore); // 85
console.log('Fraud risk:', result.fraudRisk); // 'low'
}Enterprise / ARIA V2 — embedding-first authEnterprise
Send a pre-computed 512-dim vector instead of raw audio. Keeps auth inside the 52–68 ms Phase 04 budget (target: ≤16 ms on VoiceIDVault's side).
// ARIA V2 — send a pre-computed 512-dim embedding instead of raw audio
// Skips the DSP pipeline on VoiceIDVault's side, keeping auth within the
// 52–68 ms Phase 04 budget (target: ≤16 ms on our side).
const response = await fetch(
'https://voiceidvault.com/api/v1/aria/authenticate',
{
method: 'POST',
headers: {
'Authorization': 'Bearer viv_your_api_key',
'Content-Type': 'application/json',
},
body: JSON.stringify({
payload_version: 'v2',
embedding: {
// 512-dim L2-normalised unit vector from shared-voice (:3050)
vector: Array.from(float32Array), // Float32Array → plain number[]
model_id: 'shared-voice-mfcc-v1.2.3',
dims: 512,
extraction_confidence: 0.94,
extraction_ms: 43,
sample_rate_hz: 16000,
window_ms: 2000,
deepfake: {
is_safe: true,
confidence: 0.91,
source: 'resemble',
checked_at: new Date().toISOString(), // must be <30 s old
},
},
user_id: 'user_123',
org_id: 'org_456',
}),
}
);
const { verified, session_token, timing_ms } = await response.json();
// verified: true | false
// session_token: short-lived JWT for downstream pillar access
// timing_ms: actual comparison time on VoiceIDVault side (target ≤16)Free tier available — no credit card required
Quick Start in Any Language
Get up and running with just a few lines of code
import { VoiceIDVault } from '@voiceidvault/sdk';
// Initialize the SDK
const vault = new VoiceIDVault({
apiKey: 'your-api-key',
environment: 'production'
});
// Verify a voice
const result = await vault.verify({
audioData: audioBlob,
userId: 'user-123'
});
console.log('Verification result:', result);Available for Every Stack
Native SDKs maintained by our team with weekly updates
Core Capabilities
Developer Experience
Start Building Today
Join thousands of developers using VoiceIDVault to secure their applications