Fish Audio Text-to-Speech

Overview

FishAudioTTSService provides real-time text-to-speech synthesis through Fish Audio’s WebSocket-based streaming API. The service offers custom voice models, prosody controls, and multiple audio formats optimized for conversational AI applications with low latency.

Fish TTS API Reference

Pipecat’s API methods for Fish Audio TTS integration

Example Implementation

Complete example with custom voice model

Fish Audio Documentation

Official Fish Audio documentation

Voice Models

Create and manage custom voice models

Installation

To use Fish Audio services, install the required dependencies:

uv add "pipecat-ai[fish]"

Prerequisites

Fish Audio Account Setup

Before using Fish Audio TTS services, you need:

Fish Audio Account: Sign up at Fish Audio Console
API Key: Generate an API key from your account dashboard
Voice Models: Create or select custom voice models for synthesis

Required Environment Variables

FISH_API_KEY: Your Fish Audio API key for authentication

Configuration

FishAudioTTSService

str

required

Fish Audio API key for authentication.

str

default:"None"

deprecated

Reference ID of the voice model to use for synthesis. Deprecated in v0.0.105. Use settings=FishAudioTTSService.Settings(voice=...) instead.

str

default:"None"

deprecated

Deprecated since 0.0.105, removed in 2.0.0. Use settings=FishAudioTTSService.Settings(...) instead.

str

default:"pcm"

Audio output format. Options: "pcm", "opus", "mp3", "wav". Audio is pushed unchanged in TTSAudioRawFrames, which transports play as raw PCM, so use "pcm" unless you decode the audio yourself.

int

default:"None"

Output audio sample rate in Hz. When None, uses the pipeline’s configured sample rate.

InputParams

default:"None"

deprecated

Deprecated since 0.0.105, removed in 2.0.0. Use settings=FishAudioTTSService.Settings(...) instead.

FishAudioTTSService.Settings

default:"None"

Runtime-configurable settings. See Settings below.

Settings

Runtime-configurable settings passed via the settings constructor argument using FishAudioTTSService.Settings(...). These can be updated mid-conversation with TTSUpdateSettingsFrame. See Service Settings for details.

Parameter	Type	Default	Description
`model`	`str`	`"s2.1-pro"`	Model identifier. (Inherited.)
`voice`	`str`	`None`	Voice identifier. (Inherited.)
`language`	`Language \| str`	`None`	Language used to split text into sentences; not sent to Fish Audio. (Inherited.)
`latency`	`str`	`"balanced"`	Latency mode setting.
`normalize`	`bool`	`True`	Whether to normalize audio.
`temperature`	`float`	`None`	Temperature for sampling.
`top_p`	`float`	`None`	Top-p sampling parameter.
`prosody_speed`	`float`	`1.0`	Prosody speed control.
`prosody_volume`	`int`	`0`	Prosody volume control.
`prosody_normalize_loudness`	`bool`	`None`	Normalize output loudness for consistent perceived volume. Applies to the S2 family and drama-3.
`chunk_length`	`int`	`None`	Text segment size for processing (100-300).
`min_chunk_length`	`int`	`None`	Minimum characters before splitting into a new chunk (0-100).
`condition_on_previous_chunks`	`bool`	`None`	Use previously generated audio as context for voice consistency across chunks.

Usage

Basic Setup

import os

from pipecat.services.fish.tts import FishAudioTTSService

tts = FishAudioTTSService(
    api_key=os.getenv("FISH_API_KEY"),
    settings=FishAudioTTSService.Settings(
        voice="your-voice-reference-id",
    ),
)

With Prosody Controls

tts = FishAudioTTSService(
    api_key=os.getenv("FISH_API_KEY"),
    settings=FishAudioTTSService.Settings(
        voice="your-voice-reference-id",
        model="s2.1-pro",
        prosody_speed=1.2,
        prosody_volume=3,
        latency="balanced",
    ),
)

Notes

voice required: You must specify either voice (preferred) or the deprecated reference_id parameter.
Settings updates reconnect: Changing any setting via TTSUpdateSettingsFrame disconnects and reconnects the WebSocket with the new configuration.

Event Handlers

Fish Audio TTS supports the standard service connection events:

Event	Description
`on_connected`	Connected to Fish Audio WebSocket
`on_disconnected`	Disconnected from Fish Audio WebSocket
`on_connection_error`	WebSocket connection error occurred

@tts.event_handler("on_connected")
async def on_connected(service):
    print("Connected to Fish Audio")

Get Started

Pipecat Server

Client SDKs

Pipecat Flows

Pipecat Cloud

CLI

Pipecat Context Hub

Fish Audio Text-to-Speech

Overview

Fish TTS API Reference

Example Implementation

Fish Audio Documentation

Voice Models

Installation

Prerequisites

Fish Audio Account Setup

Required Environment Variables

Configuration

FishAudioTTSService

Settings

Usage

Basic Setup

With Prosody Controls

Notes

Event Handlers

​Overview

Fish TTS API Reference

Example Implementation

Fish Audio Documentation

Voice Models

​Installation

​Prerequisites

​Fish Audio Account Setup

​Required Environment Variables

​Configuration

​FishAudioTTSService

​Settings

​Usage

​Basic Setup

​With Prosody Controls

​Notes

​Event Handlers

Overview

Installation

Prerequisites

Fish Audio Account Setup

Required Environment Variables

Configuration

FishAudioTTSService

Settings

Usage

Basic Setup

With Prosody Controls

Notes

Event Handlers