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

# Paragraphs & References

> Access individual paragraphs with three reference formats, context windows, and entity enrichment.

Paragraphs are the atomic content units of the Urantia Papers. Each one has a unique reference, full text, HTML rendering, optional audio, and optional entity mentions.

## Reference formats

The API accepts three formats for identifying a paragraph — all auto-detected:

| Format                  | Example   | Pattern                              |
| ----------------------- | --------- | ------------------------------------ |
| Standard reference      | `2:0.1`   | paperId:sectionId.paragraphId        |
| Paper.section.paragraph | `2.0.1`   | paperId.sectionId.paragraphId        |
| Global ID               | `1:2.0.1` | partId:paperId.sectionId.paragraphId |

Use whichever is most natural for your use case. The standard reference (`2:0.1`) is the most common in Urantia Book study.

## Get a paragraph

```bash theme={null}
# Using standard reference
curl https://api.urantia.dev/paragraphs/2:0.1

# Same paragraph, different format
curl https://api.urantia.dev/paragraphs/2.0.1
```

## Get a random paragraph

```bash theme={null}
curl https://api.urantia.dev/paragraphs/random
```

Returns a single random paragraph — great for daily quotes, inspiration widgets, or testing.

### Filter by length

Use `minLength` and `maxLength` to filter by character count. This is useful when you need paragraphs of a specific length — for example, voice recording prompts or card-sized quotes.

```bash theme={null}
# Paragraphs between 300-800 characters (good for 10-30s narration)
curl "https://api.urantia.dev/paragraphs/random?minLength=300&maxLength=800"

# Short paragraphs for quote cards
curl "https://api.urantia.dev/paragraphs/random?maxLength=200"

# Long paragraphs for deep reading
curl "https://api.urantia.dev/paragraphs/random?minLength=500"
```

| Parameter   | Type    | Description                               |
| ----------- | ------- | ----------------------------------------- |
| `minLength` | integer | Minimum character count of paragraph text |
| `maxLength` | integer | Maximum character count of paragraph text |

## Context window

Retrieve a paragraph with its surrounding context:

```bash theme={null}
# Get paragraph with 3 paragraphs before and after
curl https://api.urantia.dev/paragraphs/2:0.1/context?window=3
```

Returns `target` (the requested paragraph), `before` (preceding paragraphs), and `after` (following paragraphs). The `window` parameter accepts 1–10.

This is especially useful for RAG applications where an LLM needs surrounding context to give accurate answers.

## Entity enrichment

Add `?include=entities` to get typed entity mentions on any paragraph:

```bash theme={null}
curl https://api.urantia.dev/paragraphs/2:0.1?include=entities
```

Each entity includes `id`, `name`, and `type` (being, place, order, race, religion, or concept).

## Response shape

```json theme={null}
{
  "data": {
    "id": "1:2.0.1",
    "standardReferenceId": "2:0.1",
    "paperId": "2",
    "paperTitle": "The Nature of God",
    "sectionId": "0",
    "sectionTitle": "",
    "paragraphId": "1",
    "text": "Your Father in heaven, by endowing you with...",
    "htmlText": "<p>Your Father in heaven, by endowing you with...</p>",
    "audio": {
      "tts-1-hd": {
        "nova": { "format": "mp3", "url": "https://audio.urantia.dev/..." }
      }
    }
  }
}
```

<Card title="API Reference — Paragraphs" icon="terminal" href="/api-reference/endpoint/get-paragraph">
  See the full endpoint documentation with interactive examples.
</Card>