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

# Eleven v4 Turbo

> Generate low-latency speech and dialogue with ElevenLabs Eleven v4 Turbo, with audio tags and per-request voice cloning. See parameters and API examples.

Eleven v4 Turbo is the real-time model of the ElevenLabs Eleven v4 family, released on September 28, 2026, with a median inference latency of about 100 ms. It takes the same input as Eleven v4: one voice or a multi-speaker dialogue, audio tags like `[whispering]` and `[laughing]`, per-request voice cloning, and 70+ languages. Each request takes up to 10,000 characters and is billed on the characters spoken.

Try it in the browser: [Eleven v4 Turbo API page](https://unifically.com/models/elevenlabs-4-turbo).

## Model

```
elevenlabs/eleven-v4-turbo
```

## Parameters

Send **either** `text` + `voice` (one voice) **or** `dialogue` (several speakers), never both.

| Parameter | Type | Required | Default | Description |
| - | - | - | - | - |
| `text` | string | With `voice` | - | Text to speak, up to 10,000 characters. Supports audio tags such as `[whispering]` or `[laughing]` |
| `voice` | string or array | With `text` | - | Voice to use — see [Voice Options](#voice-options) |
| `dialogue` | array | Instead of `text` + `voice` | - | Multi-speaker conversation — see [Dialogue Object](#dialogue-object) |
| `language_code` | string | No | auto-detect | ISO 639-1 language code (e.g. `en`, `es`). Omit to detect the language from the text |
| `seed` | integer | No | - | Seed for reproducible generation (0–4294967295) |
| `output_format` | string | No | `mp3_44100_128` | Audio output format (see formats below) |
| `voice_settings` | object | No | - | `stability` and `similarity_boost` (see below) |
| `with_timestamps` | boolean | No | `false` | Text mode only. Return character alignment and word timings alongside the audio |

<Note>Eleven v4 Turbo has no `style`, `speed` or `use_speaker_boost` settings and does not support SSML. Direct the delivery with audio tags in the text instead.</Note>

## Voice Options

In text mode, `voice` accepts two forms:

| Form | Example | Meaning |
| - | - | - |
| Voice ID | `"voice": "21m00Tcm4TlvDq8ikWAM"` | An available voice from [`GET /v1/resources/elevenlabs/voices`](/api-reference/resources). |
| Sample URLs | `"voice": ["https://.../a.mp3", "https://.../b.mp3"]` | A voice cloned from these samples for this request only, then deleted. The caller hosts the samples; Unifically does not retain them. |

### On-Demand Voice Samples

* All samples must contain the same speaker.
* Up to 25 sample URLs are accepted per request.
* Sample URLs must be **direct downloads** — redirects and link shorteners are rejected.
* Supported audio formats: mp3, wav, ogg, opus, m4a, flac, webm.
* Voice IDs must come from `GET /v1/resources/elevenlabs/voices`; IDs from a separate ElevenLabs account are not available to Unifically.

### Dialogue Object

Each item in the `dialogue` array contains:

| Parameter | Type | Required | Description |
| - | - | - | - |
| `text` | string | Yes | This speaker's line. Supports audio tags such as `[excited]` |
| `voice` | string | Yes | [Voice ID](/api-reference/resources#elevenlabs) for this speaker |

Dialogue returns one audio file for the whole conversation. `with_timestamps` and sample-URL voices are text-mode only.

### voice\_settings

| Field | Type | Default | Description |
| - | - | - | - |
| `stability` | number | `0.5` | 0–1. Lower = more expressive and varied, higher = steadier delivery |
| `similarity_boost` | number | `0.75` | 0–1. How closely the output adheres to the reference voice. Very high values can sound less natural |

## Output Formats

| Format | Description |
| - | - |
| `mp3_22050_32` | MP3 22.05kHz 32kbps |
| `mp3_44100_64` | MP3 44.1kHz 64kbps |
| `mp3_44100_96` | MP3 44.1kHz 96kbps |
| `mp3_44100_128` | MP3 44.1kHz 128kbps (default) |
| `mp3_44100_192` | MP3 44.1kHz 192kbps |
| `opus_48000_32` | Opus 48kHz 32kbps |
| `opus_48000_64` | Opus 48kHz 64kbps |
| `opus_48000_96` | Opus 48kHz 96kbps |
| `opus_48000_128` | Opus 48kHz 128kbps |
| `opus_48000_192` | Opus 48kHz 192kbps |
| `pcm_16000` | Raw PCM 16kHz |
| `pcm_22050` | Raw PCM 22.05kHz |
| `pcm_24000` | Raw PCM 24kHz |
| `pcm_44100` | Raw PCM 44.1kHz |
| `pcm_48000` | Raw PCM 48kHz |
| `wav_44100` | WAV 44.1kHz |
| `ulaw_8000` | μ-law 8kHz (Twilio) |
| `alaw_8000` | A-law 8kHz |

## Resources

Get available voices from the [Resources API](/api-reference/resources):

```
GET /v1/resources/elevenlabs/voices
```

## Example — one voice with timestamps

```bash theme={null}
curl -X POST https://api.unifically.com/v1/tasks \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "elevenlabs/eleven-v4-turbo",
    "input": {
      "text": "[whispering] The door was already open... Someone had been here.",
      "voice": "21m00Tcm4TlvDq8ikWAM",
      "with_timestamps": true
    }
  }'
```

## Example — dialogue

```bash theme={null}
curl -X POST https://api.unifically.com/v1/tasks \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "elevenlabs/eleven-v4-turbo",
    "input": {
      "dialogue": [
        { "text": "[excited] You will not believe what I found in the attic.", "voice": "21m00Tcm4TlvDq8ikWAM" },
        { "text": "[laughing] Please tell me it is not another box of cables.", "voice": "AZnzlk1XvdvUeBnXmlld" }
      ],
      "voice_settings": { "stability": 0.4 }
    }
  }'
```

## Response

```json theme={null}
{
  "code": 200,
  "success": true,
  "data": {
    "task_id": "abc123def456",
    "status": "processing"
  }
}
```

## Completed Response

Poll `GET /v1/tasks/{task_id}` until `status` is `completed`:

```json theme={null}
{
  "code": 200,
  "success": true,
  "data": {
    "task_id": "abc123def456",
    "status": "completed",
    "output": {
      "audio_url": "https://files.unifically.com/audio/abc123.mp3",
      "alignment": {
        "characters": ["T", "h", "e"],
        "character_start_times_seconds": [0.0, 0.07, 0.12],
        "character_end_times_seconds": [0.07, 0.12, 0.2]
      },
      "normalized_alignment": { "...": "same shape, for the spoken (normalized) text" },
      "words": [
        { "word": "The", "start": 0.0, "end": 0.2 },
        { "word": "door", "start": 0.26, "end": 0.55 }
      ]
    }
  }
}
```

* `alignment`, `normalized_alignment` and `words` appear only when `with_timestamps` is `true` in text mode.
* Without timestamps, and for dialogue, the output contains only `audio_url`.

<Note>
  **Volume discounts** are available for almost all models. Reach out at [contact@unifically.com](mailto:contact@unifically.com) to discuss custom pricing.
</Note>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.