> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mwirelabs.com/llms.txt
> Use this file to discover all available pages before exploring further.

# TTS

> Lemka text-to-speech. One voice per language, one endpoint.

**Lemka TTS** is the commercial synthesis stack. Language is in the path. Adding a language does not change the contract.

```
POST /v1/audio/speech/{language}
```

| Path                   | Language | Status            |
| ---------------------- | -------- | ----------------- |
| `/v1/audio/speech/kha` | Khasi    | Live / on request |
| `/v1/audio/speech/grt` | Garo     | Live / on request |
| `/v1/audio/speech/trp` | Kokborok | Live / on request |
| `/v1/audio/speech/lus` | Mizo     | Live / on request |
| `/v1/audio/speech/nag` | Nagamese | Live / on request |
| `/v1/audio/speech/mni` | Meitei   | Live / on request |

`Live / on request` means the voice exists; it is enabled on your licensed image when that language is in the contract.

<Note>
  `POST /speak` with English Kokoro is a leftover Garo demo. It is not Lemka TTS.
</Note>