murmr Documentation

murmr is a text-to-speech API with VoiceDesign — describe any voice in natural language and generate speech instantly. Save voices for consistency, stream for low-latency playback, or use WebSocket for real-time voice agents.

VoiceDesign

Create any voice from a natural language description

Saved Voices

Reuse saved voices for consistent output

SSE Streaming

Low-latency streaming with Server-Sent Events

WebSocket Real-time

Bidirectional streaming for voice agents and LLM integration

Quick Start

Get your API key

Generate speech with VoiceDesign

Describe the voice you want in natural language. No presets needed.

Your first API call

curl -X POST "https://api.murmr.dev/v1/voices/design" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Hello, welcome to murmr!",
    "voice_description": "A warm, friendly female voice, clear and calm",
    "language": "English"
  }' \
  --output speech.wav

Play the audio

The response is a WAV audio file (24kHz, 16-bit PCM). Play it directly or save it for later use.

Key Features

VoiceDesign — Describe any voice in natural language: "A calm, professional narrator, calm and professional"

Saved Voices — Save voices you create for consistent reuse across all your content

10 Languages — English, Spanish, Portuguese, German, French, Italian, Chinese, Japanese, Korean, Russian

Low-Latency Streaming — SSE and WebSocket streaming for fast playback

OpenAI Compatible — Drop-in replacement for the OpenAI TTS API

Connection Methods

Choose the right connection method for your use case:

Method	Best For
HTTP Batch	Audiobooks, bulk generation
SSE Streaming	Demos, previews, mobile apps
WebSocket	Voice agents, LLM integration

Try Before Coding

Use the Voice Playground to experiment with VoiceDesign and find the perfect voice before writing any code.

Questions? Email us at support@murmr.dev