Chatterbox TTS is available as a Python package and can be installed via pip or from source. The package supports CUDA, CPU, and Apple Silicon (MPS) for inference.
Prerequisites
Chatterbox requires Python 3.10 or higher. The package was developed and tested on Python 3.11 on Debian 11 OS.
Hardware Requirements
Chatterbox TTS supports multiple hardware configurations:
- CUDA - NVIDIA GPUs with CUDA support (recommended for best performance)
- CPU - Any modern CPU (slower inference)
- MPS - Apple Silicon (M1/M2/M3) with Metal Performance Shaders
Installation Methods
pip install chatterbox-tts
Installing from source allows you to modify the code or dependencies. The versions of all dependencies are pinned in pyproject.toml to ensure consistency.
Verify Installation
After installation, verify that Chatterbox is working correctly:
import torch
from chatterbox.tts_turbo import ChatterboxTurboTTS
# Check available device
if torch.cuda.is_available():
device = "cuda"
elif torch.backends.mps.is_available():
device = "mps"
else:
device = "cpu"
print(f"Using device: {device}")
# Load the model
model = ChatterboxTurboTTS.from_pretrained(device=device)
print("Chatterbox TTS installed successfully!")
Next Steps
Now that you have Chatterbox installed, you can:
- Follow the Quick Start Guide to generate your first audio
- Explore the different models available in the Model Zoo
- Learn about voice cloning and paralinguistic tags