VoxCPM vs chatterbox
Open-source text-to-speech projects compared
VoxCPM and chatterbox are open-source Python projects for generating speech from text, with voice cloning. VoxCPM emphasizes tokenizer-free multilingual synthesis and natural-language voice design, while chatterbox offers model variants for multilingual, low-latency English, and resource-constrained use.
VoxCPM: Generate and Clone Multilingual Speech
VoxCPM is a tokenizer-free text-to-speech system for multilingual speech generation, voice design, and voice cloning. Its VoxCPM2 release targets teams and developers who need expressive speech synthesis and can support a 2B-parameter model.
chatterbox: Generate Speech with Resemble AI TTS Models
Chatterbox is Resemble AI’s open-source text-to-speech model family, with multilingual voice cloning and smaller English models for lower-latency or CPU use. It is a Python project suited to developers building speech generation into applications and workflows.
| VoxCPM | chatterbox | |
|---|---|---|
| Language | Python | Python |
| License | Apache-2.0 | MIT |
| Stars | 38.3k | 27.1k |
| Forks | 4.3k | 3.6k |
| Last analyzed | Oct 3, 2026 | Oct 21, 2026 |
Key differences
- VoxCPM2 supports 30 languages and can create a voice from a natural-language description without reference audio; chatterbox Multilingual V3 supports 23 languages, and its Turbo and Nano examples use reference audio for cloning.
- VoxCPM describes a tokenizer-free diffusion autoregressive approach and offers streaming generation and SFT or LoRA fine-tuning; chatterbox provides model variants with different language and compute targets.
- VoxCPM documents 48 kHz output and lists about 8 GB of VRAM for VoxCPM2; chatterbox includes a 350M-parameter English Turbo model and a 110M-parameter Nano variant for CPU or resource-constrained use.
- VoxCPM is licensed under Apache-2.0, while chatterbox is licensed under MIT.
- VoxCPM offers Python and CLI workflows plus serving and edge-inference options; chatterbox documents Python APIs and CUDA, CPU, and MPS device options.
- Chatterbox audio includes Resemble AI's Perth watermark; VoxCPM documents no comparable watermark feature.
Choose VoxCPM if you…
- need multilingual speech generation across its 30 supported languages.
- want to design voices from natural-language descriptions without reference audio.
- need streaming generation, fine-tuning, or documented serving and edge-inference options.
Choose chatterbox if you…
- want to choose among multilingual, lower-latency English, and CPU-oriented model variants.
- need Turbo or Nano expressive tags such as `[laugh]` and `[chuckle]`.
- want generated audio with a documented watermark and a method to check for it.
Compare the numbers side by side, or add more projects
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.