ChatterboxTTS

API reference for the ChatterboxTTS class

Overview

ChatterboxTTS is the flagship English text-to-speech model with advanced control over voice characteristics and generation quality. It supports classifier-free guidance and multiple sampling strategies for high-quality, expressive speech synthesis.

Class Signature

class ChatterboxTTS:
    def __init__(
        self,
        t3: T3,
        s3gen: S3Gen,
        ve: VoiceEncoder,
        tokenizer: EnTokenizer,
        device: str,
        conds: Conditionals = None,
    )

Parameters

t3 · T3 · required

The T3 text-to-speech tokens model instance

s3gen · S3Gen · required

The S3Gen vocoder model instance for token-to-audio conversion

ve · VoiceEncoder · required

Voice encoder for extracting speaker embeddings from reference audio

tokenizer · EnTokenizer · required

English text tokenizer instance

device · str · required

Device to run inference on (“cuda”, “cpu”, or “mps”)

conds · Conditionals

Optional pre-computed conditionals for voice and style. See Conditionals reference

Class Methods

from_pretrained()

Load the pre-trained ChatterboxTTS model from Hugging Face.

@classmethod
def from_pretrained(cls, device: str) -> 'ChatterboxTTS'

Parameters

device · str · required

Device to load the model on (“cuda”, “cpu”, or “mps”). Automatically falls back to “cpu” if MPS is not available

Returns

model · ChatterboxTTS

Initialized ChatterboxTTS model with pre-trained weights from ResembleAI/chatterbox

Example

from chatterbox import ChatterboxTTS
import torch

# Load on GPU
device = "cuda" if torch.cuda.is_available() else "cpu"
model = ChatterboxTTS.from_pretrained(device)

from_local()

Load the model from a local checkpoint directory.

@classmethod
def from_local(cls, ckpt_dir: str, device: str) -> 'ChatterboxTTS'

Parameters

ckpt_dir · str · required

Path to the directory containing model checkpoint files

device · str · required

Device to load the model on (“cuda”, “cpu”, or “mps”)

Returns

model · ChatterboxTTS

Initialized ChatterboxTTS model with weights loaded from local directory

Instance Methods

prepare_conditionals()

Prepare voice conditionals from an audio prompt for subsequent generation calls.

def prepare_conditionals(
    self,
    wav_fpath: str,
    exaggeration: float = 0.5
)

Parameters

wav_fpath · str · required

Path to the audio file to use as voice reference

exaggeration · float · default: "0.5"

Voice exaggeration level (0.0 to 1.0). Higher values produce more expressive speech

Example

# Prepare voice from reference audio
model.prepare_conditionals(
    wav_fpath="voice_sample.wav",
    exaggeration=0.5
)

generate()

Generate speech from text using the prepared voice conditionals.

def generate(
    self,
    text: str,
    repetition_penalty: float = 1.2,
    min_p: float = 0.05,
    top_p: float = 1.0,
    audio_prompt_path: str = None,
    exaggeration: float = 0.5,
    cfg_weight: float = 0.5,
    temperature: float = 0.8,
) -> torch.Tensor

Parameters

text · str · required

The text to convert to speech

repetition_penalty · float · default: "1.2"

Penalty for repeating tokens (1.0 = no penalty, higher values discourage repetition)

min_p · float · default: "0.05"

Minimum probability threshold for sampling. Filters out tokens below this probability

top_p · float · default: "1.0"

Nucleus sampling threshold (0.0 to 1.0). Only tokens with cumulative probability up to top_p are considered

audio_prompt_path · str

Optional path to audio file for voice cloning. If provided, will override existing conditionals

exaggeration · float · default: "0.5"

Voice exaggeration level (0.0 to 1.0). Higher values produce more expressive and animated speech

cfg_weight · float · default: "0.5"

Classifier-free guidance weight (0.0 to 1.0+). Higher values increase adherence to conditioning

temperature · float · default: "0.8"

Sampling temperature (higher = more random, lower = more deterministic)

Returns

audio · torch.Tensor

Generated audio waveform as a PyTorch tensor with shape [1, samples]. Sample rate is 44100 Hz (accessible via model.sr). Audio includes perceptual watermarking

Example

import torchaudio

# Generate speech with prepared conditionals
audio = model.generate(
    text="Hello, this is a demonstration of Chatterbox TTS.",
    temperature=0.8,
    cfg_weight=0.5,
    exaggeration=0.5,
    repetition_penalty=1.2,
    min_p=0.05
)

# Save to file
torchaudio.save("output.wav", audio, model.sr)

# Or generate with a new voice in one call
audio = model.generate(
    text="Hello world!",
    audio_prompt_path="new_voice.wav",
    exaggeration=0.7
)

Attributes

sr · int

Sample rate of generated audio (44100 Hz)

device · str

Device the model is running on

conds · Conditionals

Current voice conditionals used for generation

Notes

• This model supports classifier-free guidance (CFG) for improved quality control
• Generated audio is automatically watermarked using the Perth implicit watermarker
• Text is automatically normalized (capitalization, punctuation) before generation
• The exaggeration parameter can be updated on-the-fly without re-preparing conditionals