Skip to main content

Chatterbox TTS

Chatterbox is a family of three state-of-the-art, open-source text-to-speech models by Resemble AI. Built for developers who need high-quality, low-latency speech synthesis with zero-shot voice cloning capabilities.

Quick Start

Get up and running with Chatterbox in minutes

Installation

Install Chatterbox via pip or from source

Models Overview

Explore our three optimized TTS models

API Reference

Complete API documentation

Key Features

Zero-Shot Voice Cloning

Clone any voice with just 3-10 seconds of reference audio

Paralinguistic Tags

Add realism with [laugh], [chuckle], [cough], and more

23+ Languages

Multilingual support for global applications

Built-in Watermarking

Responsible AI with Perth watermarking technology

Model Zoo

Choose the right model for your application:
ModelSizeLanguagesKey FeaturesBest For
Chatterbox-Turbo350MEnglishParalinguistic tags, lower compute and VRAMZero-shot voice agents, production
Chatterbox-Multilingual500M23+Zero-shot cloning, multiple languagesGlobal applications, localization
Chatterbox500MEnglishCFG & exaggeration tuningGeneral zero-shot TTS with creative controls
Chatterbox-Turbo is our most efficient model yet. Built on a streamlined 350M parameter architecture, it delivers high-quality speech with less compute and just one decoding step (down from 10).

Getting Started

1

Install Chatterbox

Install via pip in under a minute:
pip install chatterbox-tts
2

Load a Model

Choose from Turbo, Original, or Multilingual models:
from chatterbox.tts_turbo import ChatterboxTurboTTS

model = ChatterboxTurboTTS.from_pretrained(device="cuda")
3

Generate Speech

Create natural-sounding speech with paralinguistic tags:
text = "Hi there! [chuckle] Ready to get started?"
wav = model.generate(text)

Live Demos

Try Chatterbox models online:

Chatterbox-Turbo Demo

Experience the fastest model with paralinguistic tags

Multilingual Demo

Test 23+ language support

Listen to Samples

Hear quality examples from all models

Join Discord

Connect with the community

Use Cases

Chatterbox excels at:
  • Voice Agents: Low-latency conversational AI with natural prosody
  • Narration: Audiobooks, podcasts, and content creation
  • Localization: Multi-language content for global audiences
  • Accessibility: Text-to-speech for visually impaired users
  • Game Development: Dynamic character voices and dialogue
  • Virtual Assistants: Natural-sounding voice interfaces

Community & Support

GitHub

Star the repository

Discord

Join our community

Resemble AI

Enterprise solutions
Made with ♥️ by Resemble AI. If you need production-ready TTS with sub-200ms latency and enterprise support, check out our competitively priced service.

Build docs developers (and LLMs) love