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

# ElevenLabs Speech

> Industry-leading text-to-speech with 21+ voices, multiple languages, and 7 model variants

## Overview

ElevenLabs provides the text-to-speech models in varg. Multiple model variants are available balancing quality, speed, and language support.

| Model ID                 | Quality | Speed      | Languages    | Credits | \~Cost |
| ------------------------ | ------- | ---------- | ------------ | ------- | ------ |
| `eleven_v3`              | Best    | Standard   | Multi        | 25      | \$0.25 |
| `eleven_multilingual_v2` | Great   | Standard   | 29 languages | 25      | \$0.25 |
| `turbo`                  | Good    | Fastest    | English      | 20      | \$0.20 |
| `eleven_turbo_v2`        | Good    | Fast       | English      | 20      | \$0.20 |
| `eleven_turbo_v2_5`      | Good    | Fast       | Multi        | 20      | \$0.20 |
| `eleven_flash_v2`        | Good    | Ultra-fast | English      | 20      | \$0.20 |
| `eleven_flash_v2_5`      | Good    | Ultra-fast | Multi        | 20      | \$0.20 |

## Quick start

<CodeGroup>
  ```typescript SDK theme={null}
  import { createVarg } from "vargai/ai"

  const varg = createVarg({ apiKey: process.env.VARG_API_KEY! })

  const result = await varg.speechModel("eleven_v3").generate({
    text: "Welcome to varg, the AI video generation platform.",
    voice: "rachel",
  })

  console.log(result.audio.url)
  ```

  ```bash cURL theme={null}
  curl -X POST https://api.varg.ai/v1/speech \
    -H "Authorization: Bearer $VARG_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "eleven_v3",
      "text": "Welcome to varg, the AI video generation platform.",
      "voice": "rachel"
    }'
  ```
</CodeGroup>

## Available voices

| Voice    | Gender | Style            | Best for         |
| -------- | ------ | ---------------- | ---------------- |
| `rachel` | Female | Calm, warm       | Narration        |
| `bella`  | Female | Soft, gentle     | Storytelling     |
| `domi`   | Female | Confident        | Presentations    |
| `elli`   | Female | Young, cheerful  | Social media     |
| `adam`   | Male   | Deep, warm       | Narration        |
| `josh`   | Male   | Young, energetic | Social media     |
| `sam`    | Male   | Raspy            | Character voices |
| `antoni` | Male   | Calm             | Podcasts         |
| `arnold` | Male   | Authoritative    | Announcements    |

<Tip>
  Any valid ElevenLabs `voice_id` works. The names above are convenience aliases for built-in voices. Browse more at [ElevenLabs Voice Library](https://elevenlabs.io/voice-library).
</Tip>

## Parameters

<ResponseField name="text" type="string" required>
  The text to convert to speech.
</ResponseField>

<ResponseField name="voice" type="string" default="rachel">
  Voice name or ElevenLabs voice ID.
</ResponseField>

<ResponseField name="model" type="string" default="eleven_multilingual_v2">
  Speech model variant (see table above).
</ResponseField>

<ResponseField name="provider_options" type="object">
  `stability` — 0 to 1 (default 0.5). Higher = more consistent, lower = more expressive.
  `similarity_boost` — 0 to 1 (default 0.75). How closely to match the voice.
</ResponseField>

## Choosing a model

| Scenario                | Recommended model            |
| ----------------------- | ---------------------------- |
| Best quality English    | `eleven_v3`                  |
| Multiple languages      | `eleven_multilingual_v2`     |
| Fast English narration  | `turbo` or `eleven_turbo_v2` |
| Real-time / interactive | `eleven_flash_v2_5`          |
| Budget                  | `turbo` (20 credits)         |

## Composition example

Use speech in a video composition with captions:

```tsx theme={null}
const narration = Speech({
  model: varg.speechModel("eleven_v3"),
  text: "This product will transform your workflow.",
  voice: "adam",
})

<Clip duration={5}>
  <Video model={varg.videoModel("kling-v3")} prompt="product showcase" duration={5} />
  {narration}
  <Captions source={narration} />
</Clip>
```

## Pricing

| Model                       | Credits | USD    |
| --------------------------- | ------- | ------ |
| `eleven_v3`                 | 25      | \$0.25 |
| `eleven_multilingual_v2`    | 25      | \$0.25 |
| `turbo` / `eleven_turbo_v2` | 20      | \$0.20 |
| `eleven_turbo_v2_5`         | 20      | \$0.20 |
| `eleven_flash_v2` / `v2_5`  | 20      | \$0.20 |

## Related models

<CardGroup cols={2}>
  <Card title="VEED Fabric" icon="comment" href="/models/lipsync/veed-fabric">
    Animate a portrait with generated speech.
  </Card>

  <Card title="Sync Lipsync" icon="lips" href="/models/lipsync/sync">
    Apply speech to existing video.
  </Card>

  <Card title="Whisper" icon="ear-listen" href="/models/transcription/whisper">
    Transcribe audio back to text.
  </Card>
</CardGroup>