Resemble AI · MIT
Chatterbox
Resemble AI’s voice-cloning models: Multilingual V3 covers 23 languages; Turbo is English-only and quicker.
How much memory it needs
Memory to run each model, from its model card or the size of its files. Smaller versions trade some quality for less memory.
| Model | Version | Needs | Business use |
|---|---|---|---|
| Chatterbox-Turbo | Full | 3.0 GB | Allowed |
| Chatterbox Multilingual V3 | Full | 3.2 GB | Allowed |
Which computers run it
The best model in this group for each machine.
| Computer | Best version | Fit |
|---|---|---|
| Laptop or PC, no graphics card, 32 GB | Chatterbox Multilingual V3 Full | Runs great |
| RTX 4060 · 8 GB | Chatterbox Multilingual V3 Full | Runs great |
| RTX 3060 · 12 GB | Chatterbox Multilingual V3 Full | Runs great |
| RTX 5060 Ti · 16 GB | Chatterbox Multilingual V3 Full | Runs great |
| RTX 4090 · 24 GB | Chatterbox Multilingual V3 Full | Runs great |
| RTX 5090 · 32 GB | Chatterbox Multilingual V3 Full | Runs great |
| Mac with M4 · 16 GB | Chatterbox Multilingual V3 Full | Runs great |
| Mac with M4 Pro · 48 GB | Chatterbox Multilingual V3 Full | Runs great |
| Mac with M5 Max · 128 GB | Chatterbox Multilingual V3 Full | Runs great |
| Ryzen AI Max+ 395 · 128 GB | Chatterbox Multilingual V3 Full | Runs great |
| DGX Spark · 128 GB | Chatterbox Multilingual V3 Full | Runs great |
| RTX PRO 6000 · 96 GB | Chatterbox Multilingual V3 Full | Runs great |
| Mac Studio with M5 Ultra · 256 GB | Chatterbox Multilingual V3 Full | Runs great |
| Mac Studio with M3 Ultra · 512 GB | Chatterbox Multilingual V3 Full | Runs great |
What it is good and bad at
Good at
- Voice cloning from a short sample
- MIT licence
Not good at
- Turbo is English only
Start it
- Chatterbox-Turbo: runs in chatterbox-tts (PyTorch) (also ONNX (ResembleAI/chatterbox-turbo-ONNX), MLX (mlx-community)). Model card
- Chatterbox Multilingual V3: runs in chatterbox-tts (PyTorch) (also ONNX (onnx-community/chatterbox-ONNX), MLX (mlx-community), GGUF community ports). Model card