Chatterbox resemble-ai/chatterbox

Resemble AI's family of open TTS models (Turbo, Nano, Multilingual) with zero-shot voice cloning from a reference clip.

Creator kit + Compare GitHub ↗ Install guide (official) ↗

Chatterbox at a glance

Checked 15 Oct 2026

Chatterbox is a family of text-to-speech models from Resemble AI: Turbo (350M, English, paralinguistic tags like [laugh]), Nano (110M, CPU-oriented, about 3x realtime on 8 CPU cores) and Multilingual V3 (500M, 23 languages). It clones a voice from a short reference clip. Code and weights are MIT. Every output carries an imperceptible Perth watermark. It does not do speech recognition or full voice agents.

Cite this pageRepoQuarry. “Chatterbox.” Checked 15 Oct 2026. https://repoquarry.dev/projects/chatterbox

How much of ElevenLabs does it replace?

2 of 4 jobs. Covers speech generation and zero-shot voice cloning; no speech recognition or hosted agent platform. Output is watermarked.

Speech generation Partial replacement

Voice cloning Partial replacement

Speech recognition Not covered

Voice agents Not covered

For speech recognition, voice agents, see other ElevenLabs alternatives. You may also need to build an interface, an API server and hosting yourself.

How much of Descript does it replace?

2 of 3 jobs. Voice cloning and text-to-speech only; no transcript-based audio or video editing.

Transcript-based editing Not covered

Voice cloning Partial replacement

Speech generation Partial replacement

For transcript-based editing, see other Descript alternatives. You may also need to build an interface, an API server and hosting yourself.

Facts, with sources

Every answer links to where we found it. We read the documentation; we did not install it.

Report a correction

QuestionAnswerSourceChecked

Runs on my machine? Yes, on GPU (CUDA), Apple MPS or CPU It targets on-device and CPU inference — running 3x faster than realtime on 8 CPU cores github.com · chatterbox#usage ↗ 1d ago

Code licence MIT github.com · LICENSE ↗ 1d ago

Model weights licence MIT (Chatterbox, Turbo, Nano and Multilingual weights on Hugging Face) huggingface.co · chatterbox ↗ 1d ago

Commercial use Yes, MIT code and weights (generated audio is watermarked) huggingface.co · chatterbox ↗ 1d ago

Paid services needed? None required; Resemble AI's paid hosted TTS service is offered as an optional extra If you like the model but need to scale or tune it for higher accuracy, check out our competitively priced TTS service github.com · chatterbox#readme ↗ 1d ago

Hardware Nano runs on CPU at about 3x realtime on 8 cores; no VRAM or RAM minimums stated Runs on CPU (3x realtime on 8 cores) github.com · chatterbox#-model-zoo ↗ 1d ago

Operating systems Tested on Debian 11 Linux; code shows an Apple "mps" device option; Windows not mentioned We developed and tested Chatterbox on Python 3.11 on Debian 11 OS github.com · chatterbox#installation ↗ 1d ago

Setup effort Easy, pip install chatterbox-tts github.com · chatterbox#installation ↗ 1d ago