# Audio narrative

> Turn any text into a fully hosted, embeddable audio experience — a branded player you can drop into any page with a single <script> tag. Paste your text, pick a voice, customize the look, done.

## What you can do

Audio Narrative converts written content into a rich audio experience, complete with a player you host nowhere and style everywhere.

| Capability | Detail |
| --- | --- |
| **Generate speech** | High-quality Arabic speech from any text, up to **10,000 characters**. |
| **Pick a voice** | Choose from multiple Munsit voices and models. |
| **Brand the player** | Customize colors and theme to match your site. |
| **Embed anywhere** | One `<script>` tag on any website. |
| **Manage history** | Revisit and tweak all generated narratives from your history. |

> Optimized for Arabic. For best results, make sure your input text is clean, well-punctuated Arabic. Consider running it through Tashkīl first for precise pronunciation.

## Create a narrative

Audio Narrative lives in the [Munsit dashboard](https://app.munsit.com/) — no API call needed. From the sidebar, open **Audio Narrative** and click **Create New**.

1

## Fill in page details

**Title** — the headline shown on the player (e.g. article title). **Author** — the byline beneath the title. **Text** — paste the content to read aloud (up to 10,000 characters). **Model** — `faseeh-v1-preview` (Faseeh). **Stability** — higher for a consistent, neutral delivery; lower for more expressive variation. **Speed** — 0.7× (slower) to 1.2× (faster). **Voice** — pick the one that fits your content.

2

## Customize the player

Switch to **Player Customization**: choose the **Default** or **Light** theme, then pick any hex color for the **background** and one for **text** (titles, controls, timestamps). A live preview updates in real time.

3

## Generate and embed

Click **Create**. Once ready, open the narrative from your history to preview the audio with the built-in player, tweak colors or theme further, and copy the embed code — a single `<script>` tag — into any webpage.

## The embed code

After generating a narrative, copy the embed code from the detail page and paste it into your HTML. Everything is configured through `data-*` attributes.

```
<script
  src="https://widget.munsit.com/audio-narrative@0.1.0-beta.js"
  data-audio="YOUR_AUDIO_URL"
  data-title="Your Article Title"
  data-author="Author Name"
  data-bg="#ffffff"
  data-color="#000000"
  data-theme="light"
></script>
```

| Attribute | Description |
| --- | --- |
| `data-audio` | URL of the generated audio file. |
| `data-title` | Title shown on the player. |
| `data-author` | Author byline shown on the player. |
| `data-bg` | Background color (hex). |
| `data-color` | Text and icon color (hex). |
| `data-theme` | `default` or `light`. |

## Generation parameters

Two dials control how the narrator sounds.

| Parameter | Range | Effect |
| --- | --- | --- |
| **Stability** | `0 – 100%` | Controls delivery consistency; higher = more neutral, lower = more expressive. |
| **Speed** | `0.7× – 1.2×` | Adjusts how fast the narrator speaks. |

## Go further

Prefer raw audio over a hosted player? Those voices are one API call away.

- [Synthesize speech](/text-to-speech/synthesize) — Generate the audio yourself and host it your way. — `POST /text-to-speech`

- [Browse voices](/text-to-speech/voices) — Find the narrator voice that fits your content. — `GET /voices`

- [Clone a voice](/text-to-speech/voice-cloning) — Narrate articles in your own custom voice. — `POST /voices/clone`

- [Tashkīl](/text-to-speech/tashkil) — Diacritize Fusha text before narration for precise reading. — `POST /tashkil/diacritize`