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

# Voice, music and 3D

> Text to speech, music generation and 3D models.

## Voice-over (text to speech)

```json theme={null}
{
  "modality": "audio",
  "track": "voice",
  "model_id": "fal-elevenlabs-v4",
  "prompt": "Welcome to our summer collection. Fresh colors, light fabrics, made to move.",
  "voice": "Rachel",
  "language": "en"
}
```

For voice, `prompt` is the exact text to speak. Available voices and languages are in `options.voices` and `options.languages` of [`GET /models?modality=audio&track=voice`](/api-reference/list-models).

## Music

```json theme={null}
{
  "modality": "audio",
  "track": "music",
  "model_id": "fal-elevenlabs-music",
  "prompt": "Upbeat lo-fi beat for a product video, 30 seconds"
}
```

## 3D models

```json theme={null}
{
  "modality": "3d",
  "model_id": "fal-meshy-v6",
  "prompt": "A stylized ceramic coffee mug with a wooden handle"
}
```

Image-to-3D models accept one reference image (`reference_image_urls` or `reference_asset_ids`). The finished asset's `url` points to the 3D file (GLB).


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