Metadata-Version: 2.4
Name: nuvox
Version: 0.1.0
Summary: Python SDK for the NuVox TTS evaluation API
Author-email: Nurix AI <hello@nurix.ai>
License: MIT
Project-URL: Homepage, https://nuvox.hack.nurixlabs.tech
Project-URL: Documentation, https://nuvox.hack.nurixlabs.tech/docs
Project-URL: Repository, https://github.com/nurixlabs/team-hackathon-team
Keywords: tts,speech,evaluation,metrics,nuvox
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Multimedia :: Sound/Audio :: Speech
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.9
Description-Content-Type: text/markdown
Requires-Dist: httpx>=0.25.0

# NuVox Python SDK

Python client for the [NuVox](https://nuvox.hack.nurixlabs.tech) TTS evaluation API.

## Install

```bash
pip install nuvox
```

## Quick Start

```python
from nuvox import NuVox

client = NuVox(api_key="nvx_your_key")

# Audio-only evaluation
result = client.evaluate.audio("speech.wav")
print(result.metrics.utmos.value)    # 4.15
print(result.metrics.snr.value)      # 64.3
print(result.metrics.nisqa.overall)  # 3.82

# Audio + transcript evaluation
result = client.evaluate.audio_text("speech.wav", "Hello, welcome to NuVox.")
print(result.metrics.wer.value)      # 0.05
print(result.metrics.bleu.value)     # 0.92
```

## Metrics

| Metric | Requires | Measures |
|--------|----------|----------|
| UTMOS | Audio | Predicted naturalness MOS (1-5) |
| SNR | Audio | Signal-to-noise ratio (dB) |
| NISQA | Audio | Multi-dim quality: coloration, discontinuity, noisiness, loudness |
| Pitch Range | Audio | Monotone/robotic detection (Hz std) |
| WER | Audio+Text | Word error rate |
| CER | Audio+Text | Character error rate |
| BLEU | Audio+Text | N-gram overlap score |
| Semantic Similarity | Audio+Text | Meaning preservation |
| Speaking Rate | Audio+Text | Words per minute |

## Environment Variable

Instead of passing `api_key`, set `NUVOX_API_KEY`:

```bash
export NUVOX_API_KEY=nvx_your_key
```

```python
client = NuVox()  # picks up from env
```

## License

MIT
