# ElevenLabs Multilingual v2

> Turn text into natural, expressive speech in 29 languages with ElevenLabs Multilingual v2 voices, with controls for stability, similarity, and style.

## Overview

- **Endpoint**: `https://queue.modelrunner.run/elevenlabs/tts/multilingual-v2`
- **Model ID**: `elevenlabs/tts/multilingual-v2`
- **Category**: sound
- **Kind**: inference
- **Tags**: elevenlabs, multilingual-v2, text-to-speech, tts, voice, audio

## Pricing

- **Price**: $0.0015 per output second

## Request Lifecycle

This model runs on the ModelRunner **asynchronous queue API** — a single POST does not return the output. Every call requires an `Authorization: Key $MODEL_RUNNER_KEY` header. Run three steps:

1. **Submit** — `POST https://queue.modelrunner.run/elevenlabs/tts/multilingual-v2` with a JSON body holding the input fields at the top level. The body may also include a reserved top-level `metadata` object — a flat string map (max 16 keys, key ≤64 / value ≤512 chars) stored on the request for your own tagging. It is never sent to the model; filter your request history with `GET https://queue.modelrunner.run/requests?metadata=<url-encoded JSON>` (exact key=value matches, AND-ed). The response carries request handles only (no output yet):

```json
{
  "status": "IN_QUEUE",
  "request_id": "<21-char id>",
  "status_url": "https://queue.modelrunner.run/elevenlabs/tts/multilingual-v2/requests/<request_id>/status",
  "response_url": "https://queue.modelrunner.run/elevenlabs/tts/multilingual-v2/requests/<request_id>",
  "cancel_url": "https://queue.modelrunner.run/elevenlabs/tts/multilingual-v2/requests/<request_id>/cancel"
}
```

2. **Poll status** — `GET <status_url>` until `status` is `COMPLETED`. Possible values are `IN_QUEUE`, `IN_PROGRESS`, `COMPLETED`, `FAILED`, `CANCELLED`. A `FAILED` request responds with HTTP 400 and an `error` field.

3. **Read result** — `GET <response_url>`. Returns the finished request, including the generated `output`:

```json
{ "id": "<request_id>", "status": "COMPLETED", "output": ..., "input": ... }
```

The JavaScript and Python SDKs below perform steps 2–3 for you. In any language without an SDK (Swift, Go, Kotlin, etc.) you must implement the polling loop and the final result fetch yourself — see the cURL example for the full flow.

### Input Schema

- **`text`** (`string`, _required_):
  The text to synthesize into speech.

- **`speed`** (`number`, _optional_):
  Speaking speed. 1 is normal; below 1 is slower, above 1 is faster.
  - Default: `1`
  - Range: `0.7` to `1.2`

- **`style`** (`number`, _optional_):
  Style exaggeration. Higher values produce more expressive, exaggerated delivery.
  - Default: `0`
  - Range: `0` to `1`

- **`voice`** (`VoiceEnum`, _optional_):
  The voice to speak with. The default is Rachel.
  - Default: `"Rachel"`
  - Options: `"Rachel"`, `"Aria"`, `"Roger"`, `"Sarah"`, `"Laura"`, `"Charlie"`, `"George"`, `"Callum"`, `"River"`, `"Liam"`, `"Charlotte"`, `"Alice"`, `"Matilda"`, `"Will"`, `"Jessica"`, `"Eric"`, `"Chris"`, `"Brian"`, `"Daniel"`, `"Lily"`, `"Bill"`

- **`next_text`** (`string | null`, _optional_):
  Text that comes after this chunk, used as a continuity hint when synthesizing long content in pieces.

- **`stability`** (`number`, _optional_):
  Voice stability. Lower is more variable and expressive; higher is steadier and more monotone.
  - Default: `0.5`
  - Range: `0` to `1`

- **`language_code`** (`string | null`, _optional_):
  ISO 639-1 language code (e.g. en, es, fr, de, ja) to enforce the synthesis language. Leave unset to let the model infer it.

- **`previous_text`** (`string | null`, _optional_):
  Text that comes before this chunk, used as a continuity hint when synthesizing long content in pieces.

- **`similarity_boost`** (`number`, _optional_):
  How closely the output matches the chosen voice's character.
  - Default: `0.75`
  - Range: `0` to `1`

- **`apply_text_normalization`** (`TextNormalizationEnum`, _optional_):
  Controls normalization of text such as numbers and abbreviations before synthesis. 'on' forces it, 'off' reads text exactly as written.
  - Default: `"auto"`
  - Options: `"auto"`, `"on"`, `"off"`

### Output Schema

_No `Output` schema properties are available._

## Default Example

**Input**

```json
{
  "text": "Bonjour et bienvenue. Aujourd'hui, nous explorons les capacités multilingues de ce modèle de synthèse vocale.",
  "speed": 1,
  "style": 0,
  "voice": "Charlotte",
  "stability": 0.5,
  "similarity_boost": 0.75,
  "apply_text_normalization": "auto"
}
```

**Output**

```json
"https://media.modelrunner.ai/uzJbhEeSpbphzIB6zotRO.mp3"
```
## Usage Examples

### cURL

The queue API is asynchronous: submit the request, poll `status_url` until it is `COMPLETED`, then read the result from `response_url`. Requires `jq`.

```bash
# 1. Submit the request (returns request handles, not the output)
SUBMIT=$(curl --silent --request POST \
  --url https://queue.modelrunner.run/elevenlabs/tts/multilingual-v2 \
  --header "Authorization: Key $MODEL_RUNNER_KEY" \
  --header "Content-Type: application/json" \
  --data '{
  "text": "Bonjour et bienvenue. Aujourd'\''hui, nous explorons les capacités multilingues de ce modèle de synthèse vocale.",
  "speed": 1,
  "style": 0,
  "voice": "Charlotte",
  "stability": 0.5,
  "similarity_boost": 0.75,
  "apply_text_normalization": "auto"
}')

STATUS_URL=$(echo "$SUBMIT" | jq -r '.status_url')
RESPONSE_URL=$(echo "$SUBMIT" | jq -r '.response_url')

# 2. Poll until the request leaves the queue / in-progress state
while true; do
  STATUS=$(curl --silent --url "$STATUS_URL" \
    --header "Authorization: Key $MODEL_RUNNER_KEY" | jq -r '.status')
  echo "Status: $STATUS"
  case "$STATUS" in
    COMPLETED) break ;;
    FAILED|CANCELLED) echo "Request $STATUS"; exit 1 ;;
  esac
  sleep 1
done

# 3. Read the finished request, including the generated output
curl --silent --url "$RESPONSE_URL" \
  --header "Authorization: Key $MODEL_RUNNER_KEY"
```

### JavaScript

```javascript
import { modelrunner } from "@modelrunner/client";

const result = await modelrunner.subscribe("elevenlabs/tts/multilingual-v2", {
  input: {
  "text": "Bonjour et bienvenue. Aujourd'hui, nous explorons les capacités multilingues de ce modèle de synthèse vocale.",
  "speed": 1,
  "style": 0,
  "voice": "Charlotte",
  "stability": 0.5,
  "similarity_boost": 0.75,
  "apply_text_normalization": "auto"
}
});

console.log(result.data);
```

### Python

```python
import asyncio
import modelrunner_ai

async def main():
    response = await modelrunner_ai.submit_async(
        "elevenlabs/tts/multilingual-v2",
        arguments={
  "text": "Bonjour et bienvenue. Aujourd'hui, nous explorons les capacités multilingues de ce modèle de synthèse vocale.",
  "speed": 1,
  "style": 0,
  "voice": "Charlotte",
  "stability": 0.5,
  "similarity_boost": 0.75,
  "apply_text_normalization": "auto"
}
    )
    result = await response.get()
    print(result["output"])

asyncio.run(main())
```

## Additional Resources

- [Playground](https://modelrunner.ai/models/elevenlabs/tts/multilingual-v2)
- [OpenAPI Schema](https://modelrunner.ai/models/elevenlabs/tts/multilingual-v2/openapi.json)
- [LLM Instructions](https://modelrunner.ai/models/elevenlabs/tts/multilingual-v2/llms.txt)