Developer Platform

Build with VoiceMaster APIs

REST APIs, official SDKs, and WebSocket streaming for speech transcription, translation, text-to-speech, phrase matching, and conversational agents across 20+ African languages.

API Endpoints

A comprehensive REST API for every speech intelligence capability. Authenticate with a bearer token and start building in minutes.

POST/v1/speech/transcribe

Convert audio speech to text with speaker labels and timestamps. Supports 20+ African languages.

POST/v1/speech/translate

Translate spoken audio from one language to another in real time or from a recorded file.

POST/v1/tts/generate

Generate natural-sounding speech from text using authentic African voices.

POST/v1/phrase/match

Match transcribed speech against your phrase library for intent detection and command matching.

POST/v1/agent/converse

Send a message to a VoiceMaster agent and receive a conversational response.

GET/v1/phrase/library

Retrieve your organization phrase library with fingerprints and categories.

Code Examples

Get started with the phrase match API in your language of choice. Copy, paste, and run.

import { VoiceMaster } from '@voicemaster/sdk';

const vm = new VoiceMaster({ apiKey: process.env.VM_API_KEY });

const result = await vm.phrase.match({
  text: 'Ete sen?',
  language: 'twi',
  threshold: 0.85,
});

console.log(result.matches);
// [{ phrase_id: '1', text: 'Ete sen?', confidence: 0.94, category: 'Greetings' }]

Official SDKs

First-class SDKs for every major language and framework. Install with a single command.

JavaScript

npm install @voicemaster/sdk

TypeScript

npm install @voicemaster/sdk

Python

pip install voicemaster

PHP

composer require voicemaster/sdk

Laravel

composer require voicemaster/laravel

Flutter

flutter pub add voicemaster

Kotlin

implementation "ai.voicemaster:sdk:1.0.0"

Java

implementation "ai.voicemaster:sdk:1.0.0"

C#

dotnet add package VoiceMaster.SDK

cURL

curl https://api.voicemaster.ai/v1/...

Real-time Streaming

WebSocket Streaming

For real-time use cases like live interpretation, meeting transcription, and conversational agents, VoiceMaster offers a WebSocket streaming API. Stream audio in chunks and receive incremental transcription, translation, and phrase match events as they happen — with sub-second latency.

Bidirectional audio streaming over WebSocket
Incremental transcription with partial results
Real-time translation between any two languages
Live phrase matching and intent detection

const ws = new WebSocket('wss://stream.voicemaster.ai/v1/speech');

ws.onopen = () => {
  ws.send(JSON.stringify({
    action: 'start',
    language: 'twi',
    translate_to: 'english',
  }));
  // stream audio chunks...
  ws.send(audioChunk);
};

ws.onmessage = (e) => {
  const msg = JSON.parse(e.data);
  // { type: 'transcript', text: 'Ete sen?', is_final: true }
  // { type: 'translation', text: 'How are you?' }
  // { type: 'phrase_match', phrase_id: '1', confidence: 0.94 }
};

Event-driven

Webhooks

Subscribe to events and receive HTTP callbacks at your endpoint. All webhooks are signed with HMAC for verification.

phrase.matched

Fired when a transcribed phrase matches a library entry.

conversation.completed

Fired when an agent conversation finishes.

harvest.completed

Fired when an audio/video harvest job finishes processing.

agent.escalated

Fired when an agent escalates a conversation to a human.

statement.confirmed

Fired when a witness confirms a recorded statement.

tts.ready

Fired when a text-to-speech generation job completes.

Start building today

Create a free account to get your API keys, or jump straight to the dashboard to manage your applications.