# Sabi TTS

> Documentation for Capella, Sabi's expressive text-to-speech voice model.

## Docs

- [Errors & limits](https://docs.sabi-tts-app.dev.neuralace.co/api-reference/errors.md): Error shapes, status codes, quotas, and rate limits.
- [Service health](https://docs.sabi-tts-app.dev.neuralace.co/api-reference/health.md): Liveness and uptime of the gateway. No authentication required.
- [List models](https://docs.sabi-tts-app.dev.neuralace.co/api-reference/models.md): Enumerate available voice models. OpenAI-compatible.
- [Generate speech](https://docs.sabi-tts-app.dev.neuralace.co/api-reference/speech.md): Synthesize a complete audio clip from text. OpenAI-compatible.
- [Stream speech](https://docs.sabi-tts-app.dev.neuralace.co/api-reference/streaming.md): Low-latency, sentence-by-sentence audio over WebSocket.
- [Audio tags](https://docs.sabi-tts-app.dev.neuralace.co/capella/audio-tags.md): Inline tokens that make Capella's speech expressive — laughs, sighs, pauses, and emotional styles.
- [Capella](https://docs.sabi-tts-app.dev.neuralace.co/capella/capella.md): Text-to-speech with Sabi's Capella voice model.
- [Introduction](https://docs.sabi-tts-app.dev.neuralace.co/introduction.md): Capella is Sabi's expressive text-to-speech voice model, served through an OpenAI-compatible API.
- [Quickstart](https://docs.sabi-tts-app.dev.neuralace.co/quickstart.md): Generate your first Capella speech clip in under five minutes.