Developers who want a voice-cloning TTS model they can run themselves get it free under MIT, with built-in watermarking.
Kenning's take
What it does
Chatterbox Turbo is a 350-million-parameter open-source text-to-speech model from Resemble AI, aimed at low-latency voice agents and narration. It can clone a voice from about five seconds of audio, supports tags such as [laugh] and [cough], and watermarks its output. Resemble now offers it free under an MIT licence and no longer sells its voice models as paid products.
- 350M parameters, up to 6x faster than real time on a GPU
- Zero-shot voice cloning from a few seconds of audio
- Paralinguistic tags like [laugh] and [cough]
- Built-in PerTh watermarking
- MIT licence, code on GitHub
Who it suits. Developers building voice agents or narration tools who want a TTS model they can run themselves.
Key facts
- Category
- AI models
- Made by
- Resemble AI
- Launched
- 2025
- How to get it
- free download from GitHub and Hugging Face; Resemble no longer sells it as a paid product
- Works on
- self-hosted, Hugging Face
- In the UK
- Yes
- Free plan
- free, MIT licence
Reviews and write-ups
- The Decoder 2025-12-27
FAQ
Is Chatterbox Turbo free?
Yes. It is open source under the MIT licence and free to build on.
Can Chatterbox Turbo clone voices?
Yes. Resemble says it clones a voice from about five seconds of reference audio, and output is watermarked.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
ACE-Step 1.5
Open-source music generation model that writes full songs with lyrics and runs on ordinary consumer GPUs.
AlphaGenome Atlas
Google DeepMind database of AI-predicted effects for every possible single-letter change in human DNA.
Amazon Nova 2
Amazon's second family of Nova AI models, covering reasoning, speech conversation and combined text-and-image generation.
Cisco Antares
Cisco's small open-weight models that find where known security vulnerabilities sit in a codebase, running locally.