Chatterbox is a text-to-speech model published on Hugging Face by Resemble AI. It generates speech audio from text input and supports voice cloning through an audio prompt.
The model handles 23 languages and is released under the MIT license. It is delivered as an open-source Python package named chatterbox-tts that can be installed via pip. Integration with the model occurs through a simple API that loads the model onto a CUDA device and produces waveform output compatible with torchaudio for saving as WAV files.
Usage begins by importing the ChatterboxTTS class and instantiating it with from_pretrained. A basic call to the generate method accepts a text string and returns audio at the model's sample rate. An optional audio_prompt_path argument allows synthesis in the voice of a provided reference WAV file. The repository supplies example code for both standard synthesis and voice-cloned synthesis, along with pointers to Colab and Kaggle notebooks for immediate testing.
Developers working on speech-generation or multilingual-tts tasks can access the model card, weights, and community contributions directly on the Hugging Face page. No pricing information appears in the listing because the model is provided as a free, open-source download under the MIT license.
Chatterbox sits in PulseGate's Voice, TTS & speech category. It enables developers to generate natural-sounding speech from text in multiple languages. Chatterbox is an open-source project aimed at speech technology developers. The project is open source (Open Source). Chatterbox is available on the command line, and it can be self-hosted.
Behind Chatterbox is Resemble AI, and the product first shipped in 2024. Among its 5 catalogued features are text-to-speech, voice cloning, and multilingual support.
Latest indexed changes and source events
ResembleAI/chatterbox verified by the PulseGate indexer
Other apps tracked under the same category.
