API Quickstart

This guide walks you through your first Extend API call. You’ll parse a document with the Parse endpoint and get back clean, LLM-ready markdown and structured blocks you can feed into RAG, downstream extraction, or any other step in your pipeline.

What we’re going to parse

We’ll parse a bank statement PDF with headers, an account summary, and a transactions table.

For this guide we’ve hosted the bank statement here.

What you’ll get back:

• Clean markdown for each page, ready to drop into an LLM prompt
• Layout-aware blocks (text, tables, figures, key-value pairs) with their positions on the page
• Page-level chunks you can index or post-process

For structured field extraction (pulling specific values like account numbers into typed JSON), see the Extract endpoint after this guide.

Get your API key

Create a key on the Developers page and store it in an environment variable:

$
export EXTEND_API_KEY="your_api_key_here"

Install the SDK

$
pip install extend-ai

Prefer raw HTTP? Use the cURL tab below and skip this step. For Maven and other install options, see the SDKs page.

Parse the document

Now let’s parse the sample document step by step. Pick your language: each tab walks through initializing the client, parsing the hosted document, and reading the result. Prefer to use your own file? Each “Parse the document” step has a tip showing how to upload it and pass the returned id instead.

1
import os
2
from extend_ai import Extend
3
4
client = Extend(token=os.environ["EXTEND_API_KEY"])

1
response = client.parse(
2
    file={"url": "https://extend-public-files.s3.us-east-2.amazonaws.com/bank_statement_example.pdf"}
3
)
4
5
print(response.status)  # PROCESSED

1
with open("bank_statement.pdf", "rb") as f:
2
    uploaded = client.files.upload(file=f)
3
4
response = client.parse(file={"id": uploaded.id})

1
for chunk in response.output.chunks:
2
    print(chunk.content)  # clean markdown for the chunk
3
    for block in chunk.blocks:
4
        print(block.type, block.content)  # text, table, figure, ...

1
import os
2
from extend_ai import Extend
3
4
client = Extend(token=os.environ["EXTEND_API_KEY"])
5
6
response = client.parse(
7
    file={"url": "https://extend-public-files.s3.us-east-2.amazonaws.com/bank_statement_example.pdf"}
8
)
9
print(response.status)
10
11
for chunk in response.output.chunks:
12
    print(chunk.content)

Understanding the response

The parsed document content lives in output.chunks. Each chunk has a content string (clean markdown for that page or section) and a blocks array of structured elements with layout coordinates.

1
{
2
  "object": "parse_run",
3
  "id": "pr_3f1j6I1gsw5k96xFiCnkM",
4
  "file": {
5
    "object": "file",
6
    "id": "file_GzKUy0VDhHscv7tweODYb",
7
    "name": "bank_statement.pdf"
8
  },
9
  "status": "PROCESSED",
10
  "output": {
11
    "chunks": [
12
      {
13
        "id": "chunk_qncr8Txe-wYvmFjipXgMD",
14
        "type": "page",
15
        "content": "CHASE JPMorgan Chase Bank, N.A. P O Box 659754...",
16
        "metadata": { "pageRange": { "start": 1, "end": 1 } },
17
        "blocks": [
18
          {
19
            "object": "block",
20
            "id": "block_WNoJ0WbMj4pRW9MpMpUox",
21
            "type": "text",
22
            "content": "CHASE JPMorgan Chase Bank, N.A. P O Box 659754 San Antonio, TX 78265 - 9754",
23
            "details": {},
24
            "metadata": { "page": { "number": 1, "width": 612, "height": 792 } },
25
            "polygon": [
26
              { "x": 56.873, "y": 35.374 },
27
              { "x": 162.173, "y": 35.215 },
28
              { "x": 162.245, "y": 81.158 },
29
              { "x": 56.938, "y": 81.317 }
30
            ],
31
            "boundingBox": { "left": 56.873, "top": 35.215, "right": 162.245, "bottom": 81.317 }
32
          }
33
        ]
34
      }
35
    ]
36
  },
37
  "metrics": { "pageCount": 7, "processingTimeMs": 8293 },
38
  "usage": { "credits": 14 }
39
}

Key fields:

Pass content straight into an LLM, or walk blocks when you need tables and spatial structure. For the complete shape, see Response Format.

Configuring the parser

The default settings work well for most documents, but you can shape the output by passing a config object. The most common options are the target format, the chunking strategy, and per-block options:

1
response = client.parse(
2
    file={
3
        "url": "https://extend-public-files.s3.us-east-2.amazonaws.com/bank_statement_example.pdf",
4
    },
5
    config={
6
        # "markdown" (default) for clean LLM-ready text, or "spatial" to preserve layout
7
        "target": "markdown",
8
        # Group content into logical sections instead of one chunk per page
9
        "chunkingStrategy": {
10
            "type": "section",
11
            "options": {"maxCharacters": 2000},
12
        },
13
        # Analyze and summarize charts, diagrams, and images
14
        "blockOptions": {"figures": {"enabled": True}},
15
    },
16
)

What these options do:

• target: "markdown" (default) returns clean, reading-order markdown ideal for LLMs and RAG. "spatial" preserves the original layout for messy scans and some extraction use cases.
• chunkingStrategy.type: "page" (default), "section" (logical, heading-aware chunks for RAG), or "document" (one chunk for the whole file).
• blockOptions: Fine-grained control over how figures, tables, and other block types are detected and formatted.

For the full list of options, see Configuration Options. You can also configure the parser visually and export the config from Extend Studio.

Next steps

You’ve parsed a document with the API. Here’s where to go next: