MiniMaxTTS

Use MiniMaxTTS when you want expressive speech synthesis from MiniMax’s Speech models via a simple REST call. Each synthesize() request returns the complete audio as a Blob — no WebSocket management required. MiniMax offers 300+ system voices across 30+ languages plus cloned voices, all selected through the voiceId option, with emotion, speed, volume, and pitch controls.

Prerequisites

• A MiniMax API key or a CompositeVoice proxy server
• No additional dependencies required. MiniMaxTTS uses native fetch internally.

Basic setup

import {
  CompositeVoice,
  NativeSTT,
  AnthropicLLM,
  MiniMaxTTS,
  BrowserAudioOutput,
} from 'composite-voice';

const voice = new CompositeVoice({
  providers: [
    new NativeSTT(),
    new AnthropicLLM({
      proxyUrl: '/api/proxy/anthropic',
      model: 'claude-haiku-4-5',
    }),
    new MiniMaxTTS({
      proxyUrl: '/api/proxy/minimax',
      voiceId: 'English_expressive_narrator',
      model: 'speech-02-hd',
      audioFormat: 'mp3',
    }),
    new BrowserAudioOutput(),
  ],
});

await voice.initialize();
await voice.startListening();

Configuration options

Available voices

MiniMax ships 300+ system voices such as English_expressive_narrator and English_Graceful_Lady. List them with the platform’s POST /v1/get_voice endpoint, or clone your own via MiniMax’s voice cloning API and pass the resulting voice ID.

Output formats

Complete example

import {
  CompositeVoice,
  MicrophoneInput,
  DeepgramSTT,
  AnthropicLLM,
  MiniMaxTTS,
  BrowserAudioOutput,
} from 'composite-voice';

const tts = new MiniMaxTTS({
  proxyUrl: '/api/proxy/minimax',
  voiceId: 'English_expressive_narrator',
  model: 'speech-02-hd',
  audioFormat: 'mp3',
  emotion: 'calm',
  speed: 1.1,
});

const voice = new CompositeVoice({
  providers: [
    new MicrophoneInput(),
    new DeepgramSTT({ proxyUrl: '/api/proxy/deepgram' }),
    new AnthropicLLM({
      proxyUrl: '/api/proxy/anthropic',
      model: 'claude-haiku-4-5',
    }),
    tts,
    new BrowserAudioOutput(),
  ],
});

voice.on('tts.start', () => console.log('Speaking...'));
voice.on('tts.end', () => console.log('Done speaking'));

await voice.initialize();
await voice.startListening();

Model selection

• speech-02-hd / speech-02-turbo — Widely deployed Speech-02 generation. hd prioritizes audio quality (default); turbo prioritizes latency.
• speech-2.6-hd / speech-2.6-turbo — Speech 2.6 generation with broader language coverage.
• speech-2.8-hd / speech-2.8-turbo — Latest generation with the best naturalness.
• speech-01-hd / speech-01-turbo — Previous generation, kept for compatibility.

Group ID

Older MiniMax API keys are scoped to a group and require a GroupId query parameter on every request; newer keys embed the group and work without it. If your requests fail with an authentication error, set the groupId option to the Group ID shown in your MiniMax console — MiniMaxTTS appends it as ?GroupId=<id>. The query parameter passes through the CompositeVoice proxy unchanged, so groupId works in proxy mode too.

Tips

• The default base URL is the global endpoint (https://api.minimax.io). Use the endpoint option to target a regional host such as https://api-uw.minimax.io (western US).
• MiniMaxTTS is REST-based, not streaming. The full audio Blob is returned after the API processes the entire input. For real-time streaming, consider DeepgramTTS.
• The MiniMax API returns audio hex-encoded; MiniMaxTTS decodes it into a playable Blob for you.
• Use proxyUrl in production so your API key stays server-side — set minimaxApiKey in your proxy config.

Further reading

• API reference
• Providers reference
• Getting started