SonexLabs Text-to-Speech

SonexTTSService converts text to speech with SonexLabs’ streaming TTS API.

Overview

SonexTTSService converts text into speech using SonexLabs’ streaming text-to-speech API. It streams audio over HTTP with connection keep-alive for low time-to-first-audio, and supports multiple voices and languages, with automatic language detection when none is specified.

Source Repository

Source code, examples, and issues for the SonexLabs integration

SonexLabs

Learn more about SonexLabs’ text-to-speech service

Installation

This is a community-maintained package distributed separately from pipecat-ai:

uv add pipecat-sonex

Prerequisites

SonexLabs Account Setup

Before using the SonexLabs text-to-speech service, you need:

1. SonexLabs Account: Sign up at SonexLabs
2. API Key: Generate an API key from your account dashboard
3. Voice Selection: Pick a voice ID from your account’s available voices

Required Environment Variables

• SONEX_API_KEY: Your SonexLabs API key for authentication

Configuration

api_key · str · required

SonexLabs API key.

voice · str · required

Voice ID to use for synthesis.

language · str · default: "None"

Optional language override. Leave unset to auto-detect from the input text.

settings · SonexTTSService.Settings · default: "None"

Runtime-configurable settings, such as speed.

Usage

from pipecat_sonex import SonexTTSService

tts = SonexTTSService(
    api_key=os.getenv("SONEX_API_KEY"),
    voice=os.getenv("SONEX_VOICE_ID"),
)

pipeline = Pipeline([
    transport.input(),
    stt,
    context_aggregator.user(),
    llm,
    tts,
    transport.output(),
    context_aggregator.assistant(),
])

Compatibility

Tested with Pipecat v1.7.0.