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

# Processing types

> How auto_segment, manual_segment, and audio_upload generate audio

The `type` property on a content item determines how BeyondWords processes it. Set `type` when you [create](/api-reference/content/create) or [update](/api-reference/content/update) an item.

## auto\_segment

Recommended for most publishers. Submit HTML for `body` (and optionally `title` and `summary`). BeyondWords segments the body automatically and generates audio asynchronously.

```json theme={null}
{
  "type": "auto_segment",
  "title": "<h1>Article title</h1>",
  "summary": "<h2>Summary</h2>",
  "body": "<p>First paragraph.</p><p>Second paragraph.</p>",
  "source_id": "cms-article-123",
  "source_url": "https://example.com/article",
  "published": true
}
```

The initial response returns `status: "queued"` and an empty `segments` array. After processing, call [Get content](/api-reference/content/show) with `?segments=full` to retrieve segment markers and timings.

### Pause markers

Add verbal pauses in HTML:

```html theme={null}
<p>The policy reduces emissions. <span data-beyondwords-pause="1.0"></span> In practice, it may do the opposite.</p>
```

Pause values are in seconds (max 3), with up to one decimal place.

## manual\_segment

For editorial tools or fine-grained control. Do **not** send `title`, `summary`, or `body`. Instead, send a `segments` array.

| Property       | Options                                                    |
| -------------- | ---------------------------------------------------------- |
| `section`      | `title`, `summary`, or `body`                              |
| `content_type` | `text` (TTS), `audio` (upload URL), or `image` (for video) |

```json theme={null}
{
  "type": "manual_segment",
  "source_id": "cms-article-123",
  "segments": [
    { "section": "title", "content_type": "text", "text": "Article title" },
    { "section": "body", "content_type": "text", "text": "First paragraph.", "marker": "para-1" },
    { "section": "body", "content_type": "text", "text": "Second paragraph.", "marker": "para-2" }
  ]
}
```

Use `marker` on text segments when you need [segment-level playback](/docs-and-guides/distribution/player/developer-guides/segment-detection) in the player.

For pauses in plain text, use `<pause time="1.0s" />`.

## audio\_upload

For podcasts or pre-recorded audio. Submit a `title`, `body` (shown in the player or feeds as metadata), and one audio segment:

```json theme={null}
{
  "type": "audio_upload",
  "title": "Episode title",
  "body": "Episode description",
  "source_id": "episode-42",
  "segments": [
    { "section": "title", "content_type": "audio", "audio_url": "https://example.com/episode.mp3" }
  ],
  "published": true
}
```

The title and body are not converted to speech. Use [Upload file](/api-reference/content/upload) if you need to upload the audio file to BeyondWords first. To replace AI-generated audio on an existing item, see [Replace audio](/api-reference/content/replace-audio).

## Choosing a type

| Type             | Best for                                                   |
| ---------------- | ---------------------------------------------------------- |
| `auto_segment`   | CMS integrations, article-to-audio pipelines               |
| `manual_segment` | Custom editorial UIs, mixed media, precise segment control |
| `audio_upload`   | Podcasts, human-recorded narration                         |