Get elements

The Get elements endpoint returns the parsed elements (chunks/partitions) of a source in the same format as Get build status elements. Each item includes explicit fields: element_id, element_type, text, markdown, html, img_base64 (optional), position, page_number, bounding_box, page_layout, and more.

Endpoint overview

HTTP Method

GET

Endpoint URL

https://sources.graphorlm.com/get-elements

Authentication

This endpoint requires authentication using an API token. Include your API token as a Bearer token in the Authorization header.

Learn how to create and manage API tokens in the API Tokens guide.

Request format

Headers

Header	Value	Required
`Authorization`	`Bearer YOUR_API_TOKEN`	Yes

Query parameters

Parameter	Type	Required	Description
`file_id`	string	Yes	Unique identifier of the source
`page`	integer	No	1-based page number. Use with `page_size` to enable pagination
`page_size`	integer	No	Number of elements per page (1–100). Use with `page`
`suppress_img_base64`	boolean	No	When `true`, `img_base64` is omitted from each element (reduces payload size)
`type`	string	No	Filter by element type (e.g. `NarrativeText`, `Title`, `Table`)
`page_numbers`	list of integers	No	Restrict to specific page numbers (repeat param for multiple: `?page_numbers=1&page_numbers=2`)
`elementsToRemove`	list of strings	No	Element types to exclude (repeat param for multiple)

When page and page_size are omitted, all elements are returned (no pagination). When only one is provided, pagination is not applied.

Response format

Success response (200 OK)

Paginated response with items as BuildStatusElement (same shape as elements in Get build status):

{
  "items": [
    {
      "element_id": "0ee55f099828817da5485796b339aeab",
      "element_type": "Title",
      "text": "Attention Is All You Need",
      "markdown": "## Attention Is All You Need",
      "html": "<h2>Attention Is All You Need</h2>",
      "img_base64": null,
      "position": 5,
      "page_number": 1,
      "bounding_box": {
        "left": 211.488,
        "top": 148.43,
        "width": 188.41,
        "height": 17.22
      },
      "page_layout": { "width": 612.0, "height": 792.0 },
      "page_annotation": null,
      "page_keywords": null,
      "page_topics": null,
      "metadata": {}
    }
  ],
  "total": 393,
  "page": 1,
  "page_size": 10,
  "total_pages": 40
}

Pagination fields

Field	Type	Description
`items`	array	Elements in the current page (or all elements if pagination not used)
`total`	integer	Total number of elements (matching filters)
`page`	integer \| null	Current page (1-based), or null when no pagination
`page_size`	integer \| null	Elements per page, or null when no pagination
`total_pages`	integer \| null	Total pages, or null when no pagination

Element fields (BuildStatusElement)

Field	Type	Description
`element_id`	string \| null	Unique identifier for the element
`element_type`	string \| null	Type: e.g. `Title`, `NarrativeText`, `Table`, `Image`
`text`	string	Plain text content
`markdown`	string \| null	Markdown representation when available
`html`	string \| null	HTML representation when available
`img_base64`	string \| null	Base64-encoded image data (omitted when `suppress_img_base64=true`)
`position`	integer \| null	Order/position within the document
`page_number`	integer \| null	Page number (1-based) where the element appears
`bounding_box`	object \| null	Bounding box (e.g. left, top, width, height) when available
`page_layout`	object \| null	Page dimensions (width, height) when available
`page_annotation`	string \| null	Annotation/summary for the page
`page_keywords`	array \| null	Keywords extracted for the page
`page_topics`	array \| null	Topics extracted for the page
`metadata`	object	Additional metadata

Element types

Type	Description
`Title`	Document and section titles
`NarrativeText`	Main body paragraphs
`ListItem`	Bullet or numbered list items
`Table`	Data tables
`TableRow`	Rows within tables
`Image`	Pictures or graphics
`Header`	Header content
`Footer`	Footer content
`Formula`	Mathematical formulas
`FigureCaption`	Captions for figures
`PageNumber`	Page numbering
`CodeSnippet`	Code segments
`Link`	Hyperlinks
`UncategorizedText`	Text that doesn’t fit other categories

Code examples

JavaScript/Node.js

const getElements = async (apiToken, fileId, options = {}) => {
  const params = new URLSearchParams({ file_id: fileId });
  if (options.page != null) params.set("page", String(options.page));
  if (options.pageSize != null) params.set("page_size", String(options.pageSize));
  if (options.suppressImgBase64) params.set("suppress_img_base64", "true");
  if (options.type) params.set("type", options.type);
  if (options.pageNumbers?.length) options.pageNumbers.forEach((n) => params.append("page_numbers", String(n)));
  if (options.elementsToRemove?.length) options.elementsToRemove.forEach((t) => params.append("elementsToRemove", t));

  const url = `https://sources.graphorlm.com/get-elements?${params}`;
  const response = await fetch(url, {
    method: "GET",
    headers: { Authorization: `Bearer ${apiToken}` },
  });
  if (!response.ok) throw new Error(`Failed: ${response.status} ${await response.text()}`);
  return response.json();
};

// Usage: first page of titles
getElements("grlm_your_api_token_here", "file_abc123", { page: 1, pageSize: 10, type: "Title" })
  .then((data) => data.items.forEach((el) => console.log(`${el.element_type}: ${el.text}`)))
  .catch(console.error);

Python

import requests

def get_elements(api_token, file_id, page=None, page_size=None, suppress_img_base64=False,
                 element_type=None, page_numbers=None, elements_to_remove=None):
    url = "https://sources.graphorlm.com/get-elements"
    headers = {"Authorization": f"Bearer {api_token}"}
    params = {"file_id": file_id}
    if page is not None:
        params["page"] = page
    if page_size is not None:
        params["page_size"] = min(100, max(1, page_size))
    if suppress_img_base64:
        params["suppress_img_base64"] = "true"
    if element_type:
        params["type"] = element_type
    if page_numbers:
        params["page_numbers"] = page_numbers
    if elements_to_remove:
        params["elementsToRemove"] = elements_to_remove

    response = requests.get(url, headers=headers, params=params, timeout=30)
    response.raise_for_status()
    return response.json()

# Usage: tables from pages 2–4
data = get_elements(
    "grlm_your_api_token_here",
    "file_abc123",
    page=1,
    page_size=50,
    element_type="Table",
    page_numbers=[2, 3, 4],
)
for el in data["items"]:
    print(f"Page {el['page_number']}: {el['text'][:100]}...")

cURL

# First page, 10 elements
curl -X GET "https://sources.graphorlm.com/get-elements?file_id=file_abc123&page=1&page_size=10" \
  -H "Authorization: Bearer grlm_your_api_token_here"

# Only NarrativeText, exclude images from payload
curl -X GET "https://sources.graphorlm.com/get-elements?file_id=file_abc123&type=NarrativeText&suppress_img_base64=true" \
  -H "Authorization: Bearer grlm_your_api_token_here"

# Filter by page numbers and exclude some types
curl -X GET "https://sources.graphorlm.com/get-elements?file_id=file_abc123&page_numbers=1&page_numbers=2&elementsToRemove=PageNumber&elementsToRemove=Footer" \
  -H "Authorization: Bearer grlm_your_api_token_here"

Error responses

Status code	Description
`400`	Invalid input (e.g. missing `file_id`)
`404`	Source file not found
`500`	Internal server error while loading elements

Example body:

{ "detail": "file_id is required" }

{ "detail": "File not found" }

{ "detail": "Internal server error occurred while loading file elements" }

Best practices

Use file_id: Obtain it from List sources or Get build status after ingestion.
Reduce payload: Set suppress_img_base64=true when you don’t need image data.
Filter server-side: Use type, page_numbers, and elementsToRemove to limit results.
Pagination: Use page and page_size (max 100) for large documents to avoid large responses.

Next steps

Get build status

Poll build status and optionally get elements for an async ingestion

List sources

List all sources and their file_ids

Upload sources

Ingest files, URLs, GitHub, or YouTube (async)

Reprocess source

Re-process a source with a different partition method (async)

Get Started

Data API Options

RAG Pipelines API

Endpoint overview

HTTP Method

Endpoint URL

Authentication

Request format

Headers

Query parameters

Response format

Success response (200 OK)

Element fields (BuildStatusElement)

Element types

Code examples

JavaScript/Node.js

Python

cURL

Error responses

Best practices

Next steps

Get build status

List sources

Upload sources

Reprocess source

Get Started

Data API Options

RAG Pipelines API

​Endpoint overview

HTTP Method

Endpoint URL

​Authentication

​Request format

​Headers

​Query parameters

​Response format

​Success response (200 OK)

​Pagination fields

​Element fields (BuildStatusElement)

​Element types

​Code examples

​JavaScript/Node.js

​Python

​cURL

​Error responses

​Best practices

​Next steps

Get build status

List sources

Upload sources

Reprocess source

Endpoint overview

Authentication

Request format

Headers

Query parameters

Response format

Success response (200 OK)

Pagination fields

Element fields (BuildStatusElement)

Element types

Code examples

JavaScript/Node.js

Python

cURL

Error responses

Best practices

Next steps