# ElevenLabs

> Expressive, cloneable voices. Every speaking setting for ElevenLabs, and how to set it.

Source: https://voxio.in/docs/voice-speaking-elevenlabs · voxioagent 2.0.0 · Index: https://voxio.in/docs/llms.txt

Service: `elevenlabs`. A new setup starts from:

```json
{
  "service": "elevenlabs",
  "voice": "",
  "model": "eleven_flash_v2_5"
}
```

## Set it

On the flow's **Voice** tab, choose **Speaking**, then **ElevenLabs**. Or from code:

```js
await voxio.flows.update(flowKey, {
  tts: {"service":"elevenlabs","voice":"","model":"eleven_flash_v2_5"},
});
```

## Settings

Settings marked “under More settings” are behind **More settings** in the dashboard.

- `voice` (string): **Voice ID**. Copy it from your ElevenLabs voice library.
- `model` (string): **Model**. Which voice model speaks. Hover an option to compare. Default: `"eleven_flash_v2_5"`.
- `language` (string): **Language** (under More settings). Force a language instead of detecting it.
- `speed` (number (×)): **Speed**. How fast the agent talks. From 0.7 to 1.2. Default: `1`.
- `stability` (number): **Stability**. Higher is steadier; lower is more varied. From 0 to 1. Default: `0.5`.
- `similarity-boost` (number): **Voice likeness** (under More settings). How closely it matches the original voice. From 0 to 1. Default: `0.75`.
- `style` (number): **Style** (under More settings). Exaggerates the voice style. Keep low for calls. From 0 to 1. Default: `0`.
- `use-speaker-boost` (boolean): **Speaker boost** (under More settings). Sharper, more present voice. Default: `true`.
- `apply-text-normalization` (string): **Normalise text** (under More settings). Read numbers and abbreviations naturally. Default: `"auto"`.

## Choices

### Model choices

| Value | Name | Notes |
| --- | --- | --- |
| `eleven_flash_v2_5` | Flash v2.5 | Lowest latency. Starts speaking fastest. The best choice for live calls. |
| `eleven_turbo_v2_5` | Turbo v2.5 | A little richer than Flash, slightly slower. |
| `eleven_multilingual_v2` | Multilingual v2 | Highest quality. Most natural, but slower to start. Better for recordings than live calls. |
| `eleven_v3` | v3 | Most expressive. Emotional and dramatic delivery. Slowest to start. |

### Normalise text choices

| Value | Name | Notes |
| --- | --- | --- |
| `auto` | auto |  |
| `on` | on |  |
| `off` | off |  |

## Next steps

- [Speaking](https://voxio.in/docs/voice-speaking.md): compare the other providers.
- [Turn-taking](https://voxio.in/docs/turn-taking.md): when the agent answers.