Metadata-Version: 2.4
Name: vtts
Version: 0.1.2
Summary: Unified TTS wrapper — kspeak (Kokoro 82M) and cspeak (Chatterbox Turbo) in one simple API.
Author: Vaibhav Upadhyaya
License-Expression: MIT
Project-URL: Homepage, https://pypi.org/project/vtts/
Keywords: tts,text-to-speech,kokoro,chatterbox,voice-cloning
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Multimedia :: Sound/Audio :: Speech
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: soundfile
Provides-Extra: kokoro
Requires-Dist: kokoro; extra == "kokoro"
Provides-Extra: chatterbox
Requires-Dist: chatterbox-tts; extra == "chatterbox"
Provides-Extra: all
Requires-Dist: kokoro; extra == "all"
Requires-Dist: chatterbox-tts; extra == "all"
Dynamic: license-file

# vtts — Voice TTS

Unified TTS wrapper for **Kokoro 82M** and **Chatterbox Turbo** in one simple API.

## Installation

```bash
# Base install
pip install vtts

# With Kokoro support
pip install vtts[kokoro]

# With Chatterbox support
pip install vtts[chatterbox]

# Everything
pip install vtts[all]
```

## Usage

### Kokoro 82M — `kspeak`

```python
from vtts import kspeak

kspeak("Hello world!")                                    # defaults: voice="af_heart"
kspeak("Different voice.", voice="af_bella")              # change voice
kspeak("Save somewhere.", output="greet.wav")             # custom output
kspeak("Speak faster.", speed=1.3)                        # adjust speed
```

### Chatterbox Turbo — `cspeak`

```python
from vtts import cspeak

cspeak("Hello world!", ref="voice.wav")                   # requires a reference voice WAV
cspeak("Clone a voice.", ref="other_voice.wav")           # different reference
cspeak("Ha! [laugh] That's funny.", ref="voice.wav")      # paralinguistic tags
```

## Notes

- **Kokoro 82M** is lightweight (82M params) and runs on CPU. Uses `espeak-ng` for phonemization.
- **Chatterbox Turbo** is larger and runs on CUDA GPU. Supports zero-shot voice cloning.
- Models are loaded lazily on first call and cached — subsequent calls are fast.
