# Text to Speech

Configure and use text-to-speech voice output for your stream.

Source: https://docs.domdimabot.com/tts/
Language: en
Documentation index: https://docs.domdimabot.com/llms.txt

Text-to-speech (TTS) reads viewer messages aloud during your stream. DomDimaBot supports two TTS providers with different voice qualities, costs, and capabilities.

## Providers

### Piper — Local (All Plans)

Piper is a local neural voice engine running on the DomDimaBot server. It is available on all plan tiers and uses AI credits.

**AI Credit cost:** 1 credit per 50 characters

**Voices available:**

* English (US) — `en_US-ryan-medium`
* Spanish (Mexico) — `es_MX-ald-medium`

**Characteristics:**

* Fast synthesis (local, no API calls)
* No expressive/emotion tags
* No premium AI features
* Available to all users regardless of plan tier

***

### Fish Audio — Cloned Voices (All Plans)

Fish Audio provides cloned character voices. It is available on all plan tiers and uses AI credits.

**Voices available:**

* Gojo, Rias Gremory, Carlos Bodoque, Toji Fushiguro

**Characteristics:**

* Cloned character voices
* Available through `$(tts.fish ...)` and `$(tts.clone ...)`
* Also usable as the channel default engine

***

## AI Credit Costs Summary

| Provider   | AI Credits          |
| ---------- | ------------------- |
| Piper      | 1 per 50 characters |
| Fish Audio | Uses AI credits     |

## Monthly AI Credits by Plan

| Plan    | Monthly Credits |
| ------- | --------------- |
| Free    | 25,000          |
| Premium | 125,000         |
| Pro     | 500,000         |

> note:
>
> AI credits are consumed when a TTS message is queued for synthesis. The credit cost is calculated on the character count of the spoken message.

## Commands

### `!s` — Speak

**Permission:** Everyone

**Usage:** `!s <message>`

Makes DomDimaBot read your message aloud using the channel's default TTS provider and voice settings.

**Example:**

```
!s Hello chat, welcome to the stream
```

### AST Functions

In custom commands and the AST system, use these functions for fine-grained control:

| Function           | Provider         | Notes                             |
| ------------------ | ---------------- | --------------------------------- |
| `$(tts ...)`       | Default provider | Uses channel's default TTS engine |
| `$(tts.speak ...)` | Piper only       | Forces the local Piper voice      |
| `$(tts.ai ...)`    | Default provider | Same as `$(tts ...)`              |
| `$(tts.clone ...)` | Fish Audio       | Cloned voice TTS                  |
| `$(tts.fish ...)`  | Fish Audio       | Same as `$(tts.clone ...)`        |

**Examples:**

```
$(tts.speak Hello from Piper)
$(tts.fish gojo Hello from Fish Audio)
```

## Channel Settings

Configure TTS in your dashboard under **Modules → Text to Speech**:

* **Enable/disable** channel TTS
* **Default provider** selection (Piper or Fish Audio)
* **Voice selection** per language or cloned voice
* **Message filters** — skip emotes, replace links, normalize whitespace, max length
* **Queue settings** — maximum pending TTS items