LandingAI Agentic Document Extraction

https://docs.landing.ai/ade
Admin
LandingAI Agentic Document Extraction (ADE) is a document intelligence platform that converts...
Tokens:87,415
Snippets:992
Trust Score:10
Update:2 months ago
Show doc for...
Tokens:
Raw
### Environment and Library Setup

Source: https://docs.landing.ai/ade/ade-quickstart

Commands to configure the API key and install the required Python library.

```bash
export VISION_AGENT_API_KEY="<your-api-key>"
```

```bash
pip install landingai-ade
```

--------------------------------

### Install the ADE Python library

Source: https://docs.landing.ai/ade/ade-python

Use pip to install the landingai-ade package.

```bash
pip install landingai-ade
```

--------------------------------

### Install jq command-line tool

Source: https://docs.landing.ai/ade/ade-quickstart

Install the jq utility to process JSON responses from the API.

```bash
brew install jq
```

```bash
winget install jqlang.jq
```

```bash
sudo apt-get install jq
```

--------------------------------

### Install the ADE TypeScript library

Source: https://docs.landing.ai/ade/ade-quickstart

Use npm to install the landingai-ade package.

```bash
npm install landingai-ade
```

--------------------------------

### Run the TypeScript script

Source: https://docs.landing.ai/ade/ade-quickstart

Execute the quickstart script using npx and view the output.

```bash
npx tsx quickstart.ts
```

```text
Extracted fields:
{ name: 'Sarah J. Mitchell', number_deposits: 5 }
```

--------------------------------

### Custom Prompts Format Example

Source: https://docs.landing.ai/ade/ade-parse-troubleshoot

The required JSON structure for the custom_prompts parameter.

```json
{"figure": "your prompt"}
```

--------------------------------

### Install Zod Dependency

Source: https://docs.landing.ai/ade/ade-typescript

Required package installation for using Zod schemas with the library.

```bash
npm install zod
```

--------------------------------

### Extract API Input Markdown Example

Source: https://docs.landing.ai/ade/ade-credit-consumption

Example of Markdown content used as input for the Extract API, totaling 2,270 characters.

```markdown
<a id='2800ae73-bd32-4c54-910d-755f25541b15'></a>

AROMA CAFE
1211 Green Street
New York, NY 10005

<a id='b7c067c1-4c96-4f07-8a2e-f3557aeb9b95'></a>

12/27/2019

<a id='9d2262d5-9319-41f1-9c69-e7ba0b726f9a'></a>

08:26 PM

<a id='8dfee4e4-c1a1-4fc8-9525-8c0352697e90'></a>

TAB 8          HOST MAGGIE

<a id='25f864af-7281-4c83-95bd-2e499dd8ee0d'></a>

AMEX ##########19883

<a id='4486463b-f5e3-487b-b9df-4959587c5091'></a>

<table>
  <tr>
    <th>QTY</th>
    <th>DESC</th>
    <th>AMT</th>
  </tr>
  <tr>
    <td>1</td>
    <td>Americano</td>
    <td>$3.19</td>
  </tr>
  <tr>
    <td>1</td>
    <td>Almond Scone</td>
    <td>$1.99</td>
  </tr>
  <tr>
    <td>1</td>
    <td>16oz Bottle Water</td>
    <td>$2.99</td>
  </tr>
</table>

<a id='91b1d921-163e-406b-bf7f-a1731f706e1c'></a>

AMT: $8.70

SUBTOTAL: $8.17
TAX: $0.53

BALANCE: $8.70

<a id='7bcc9a3e-c688-4ff2-a5b0-bfea57d25f3c'></a>

Summary : This image displays a barcode consisting of a series of vertical black and white lines, used for encoding information in a machine-readable format.

barcode:
  Barcode Details :
    • The barcode is composed of alternating black and white vertical bars of varying thickness.
    • No human-readable numeric or alphanumeric string is visible below or above the barcode.
    • Barcode type is not explicitly indicated; the pattern resembles a standard linear barcode (e.g., Code 128 or EAN-13), but cannot be definitively identified without further context.
    • The barcode contains approximately 60–70 modules (individual bars and spaces).
    • Quiet zones (blank margins) are present on both the left and right sides of the barcode.
    • The barcode is horizontally oriented and centered on a white background.
    • No additional text, logo, or annotation is present.

Analysis :
  • The barcode is designed for machine scanning and does not convey any human-readable information in this image.
  • The absence of a numeric string or label means the encoded data cannot be determined visually.
  • The presence of quiet zones and consistent bar height suggests it is formatted for standard retail or inventory use.

<a id='d06c76e1-fb44-4efb-89b7-f7eedd976062'></a>

SAMPLE RECEIPT
```

--------------------------------

### Extract API Output JSON Example

Source: https://docs.landing.ai/ade/ade-credit-consumption

Example of the JSON extraction object returned by the API, totaling 475 characters.

```json
{'merchant': {'name': 'AROMA CAFE', 'address': '1211 Green Street', 'city_state_zip': 'New York, NY 10005'}, 'transaction': {'date': '2019-12-27', 'time': '20:26', 'payment_method': 'AMEX ##########19883', 'tab_host': 'TAB 8 HOST MAGGIE'}, 'items': [{'qty': '1', 'desc': 'Americano', 'amt': '3.19'}, {'qty': '1', 'desc': 'Almond Scone', 'amt': '1.99'}, {'qty': '1', 'desc': '16oz Bottle Water', 'amt': '2.99'}], 'totals': {'subtotal': '8.17', 'tax': '0.53', 'total': '8.70'}}
```

--------------------------------

### Install ADE Skills via Claude Code

Source: https://docs.landing.ai/ade/build-with-ai-agents

Commands to add and install the ADE document processing skills plugin in Claude Code.

```bash
/plugin marketplace add landing-ai/ade-document-processing-skills
/plugin install ade-document-processing@ade-document-processing-skills
```

--------------------------------

### Root Type Schema Examples

Source: https://docs.landing.ai/ade/ade-extract-troubleshoot

Comparison between a correct schema with an object root and an incorrect schema with an array root.

```json
{
  "type": "object",
  "properties": {
    "invoice_number": {
      "type": "string",
      "description": "Invoice number"
    }
  }
}
```

```json
{
  "type": "array",
  "items": {
    "type": "object",
    "properties": {
      "invoice_number": {
        "type": "string",
        "description": "Invoice number"
      }
    }
  }
}
```

--------------------------------

### Parse and Extract Specific File

Source: https://docs.landing.ai/ade/ade-sf-extract-cloud

Concrete example of parsing a specific PDF and extracting employee payroll fields.

```sql
USE "AGENTIC_DOCUMENT_EXTRACTION__APP";

-- Step 1: Parse the staged file
CALL api.parse(
    '@DEMO_DB.DEMO_SCHEMA.DEMO_STAGE/statement-jane-harper.pdf'
);

-- Step 2: Extract structured data using inline schema
CALL api.extract(
    markdown => (SELECT MARKDOWN FROM db.parse_output WHERE FILENAME = 'statement-jane-harper.pdf'),
    schema => '{
        "title": "Employee Payroll Field Extraction Schema",
        "description": "Schema for extracting key employee payroll fields.",
        "type": "object",
        "properties": {
            "employee_name": {
                "title": "Employee Name",
                "description": "The full name of the employee as it appears on the payroll document.",
                "type": "string"
            },
            "employee_ssn": {
                "title": "Employee Social Security Number",
                "description": "The Social Security Number of the employee, formatted as XXX-XX-XXXX.",
                "type": "string"
            }
        }
    }',
    doc_id => (SELECT DOC_ID FROM db.parse_output WHERE FILENAME = 'statement-jane-harper.pdf')
);

-- Step 3: View extracted fields
SELECT
    EXTRACTION:employee_name::STRING AS employee_name,
    EXTRACTION:employee_ssn::STRING AS employee_ssn
FROM db.extract_output;
```

--------------------------------

### Document Split Configurations

Source: https://docs.landing.ai/ade/ade-json-response

Examples showing the structure of the splits array when the split parameter is omitted versus when it is set to page-level.

```json
"splits": [
    {
        "class": "full",
        "identifier": "full",
        "pages": [0],
        "markdown": "...",
        "chunks": ["chunk-id-1", "chunk-id-2"]
    }
]
```

```json
"splits": [
    {
        "class": "page",
        "identifier": "page_0",
        "pages": [0],
        "markdown": "...",
        "chunks": ["chunk-id-1"]
    },
    {
        "class": "page",
        "identifier": "page_1",
        "pages": [1],
        "markdown": "...",
        "chunks": ["chunk-id-2", "chunk-id-3"]
    }
]
```

--------------------------------

### Root Object Schema Examples

Source: https://docs.landing.ai/ade/ade-extract-troubleshoot

Comparison between a correct schema wrapped in an object and an incorrect schema provided as an array.

```json
{
  "type": "object",
  "properties": {
    "invoice_number": {"type": "string"},
    "vendor_name": {"type": "string"},
    "total": {"type": "number"}
  }
}
```

```json
["invoice_number", "vendor_name", "total"]
```

--------------------------------

### Grounding Object Example

Source: https://docs.landing.ai/ade/ade-json-response

A sample key-value pair from the grounding object representing a logo chunk.

```json
{
  "grounding": {
    "49c7b7d0-2d8e-4485-8306-3dd820eb13ed": {
      "box": {
        "left": 0.06869561225175858,
        "top": 0.02685479447245598,
        "right": 0.1445186734199524,
        "bottom": 0.08676017820835114
      },
      "page": 0,
      "type": "chunkLogo"
    }
  }
}
```

--------------------------------

### Extract Using Explicit Markdown

Source: https://docs.landing.ai/ade/ade-sf-extract-cloud

Example of querying Markdown from the parse_output table and passing it to the extraction procedure.

```sql
CALL api.extract(
    markdown => (SELECT MARKDOWN FROM db.parse_output WHERE doc_id = 'your_doc_id'),
    schema => 'https://ade.landing.ai/pdfs/InvoiceExtractionSchema.json',
    doc_id => 'your_doc_id'
);

SELECT * FROM db.extract_output;
```

--------------------------------

### Markdown Table of Contents Example

Source: https://docs.landing.ai/ade/ade-section-response

A sample Markdown-formatted table of contents string used for in-document navigation.

```text
- [Introduction](#uuid_a1)
- [Methods](#uuid_b2)
    - [Data Collection](#uuid_c3)
        - [Survey Design](#uuid_d4)
```

--------------------------------

### Visualize ADE Parse Results in TypeScript

Source: https://docs.landing.ai/ade/ade-parse-visualize-sample

Setup for drawing bounding boxes using the canvas library in a Node.js environment.

```typescript
import LandingAIADE from "landingai-ade";
import fs from "fs";
import path from "path";
import { createCanvas, loadImage } from "canvas";
import { pdf } from "pdf-to-img";

// Define colors for each chunk type
const CHUNK_TYPE_COLORS: Record<string, [number, number, number]> = {
  chunkText: [40, 167, 69],        // Green
  chunkTable: [0, 123, 255],       // Blue
  chunkMarginalia: [111, 66, 193], // Purple
  chunkFigure: [255, 0, 255],      // Magenta
  chunkLogo: [144, 238, 144],      // Light green
  chunkCard: [255, 165, 0],        // Orange
  chunkAttestation: [0, 255, 255], // Cyan
  chunkScanCode: [255, 193, 7],    // Yellow
  chunkForm: [220, 20, 60],        // Red
  tableCell: [173, 216, 230],      // Light blue
  table: [70, 130, 180],           // Steel blue
};

function rgbToString(rgb: [number, number, number]): string {
  return `rgb(${rgb[0]}, ${rgb[1]}, ${rgb[2]})`;
}
```

--------------------------------

### View Extraction Output JSON

Source: https://docs.landing.ai/ade/ade-parse-extract-sample

Example of the JSON structure generated in the output folder, containing extracted values and metadata.

```json
{
  "extraction": {
    "bank_name": "JPMorgan Chase Bank, N.A.",
    "total_invoice_amount": 15750.0
  },
  "extraction_metadata": {
    "bank_name": {
      "references": [
        "4f64f8d9-ff3a-4c47-aeb5-2ab6eaa9ce7a"
      ],
      "value": "JPMorgan Chase Bank, N.A."
    },
    "total_invoice_amount": {
      "references": [
        "deeb001e-6b3e-4c4e-96b1-6f321521ad4f",
        "0-h"
      ],
      "value": 15750.0
    }
  },
  "metadata": {
    "credit_usage": 0.5396,
    "duration_ms": 11536,
    "filename": "upload.md",
    "job_id": "bec005b58d144096b0525af3aa6ed12d",
    "org_id": null,
    "version": "extract-20260314",
    "fallback_model_version": null,
    "schema_violation_error": null,
    "warnings": []
  }
}
```

--------------------------------

### Simple Schema Metadata with Chunk IDs

Source: https://docs.landing.ai/ade/ade-extract-response

Example of metadata structure when data is extracted from text chunks using UUIDs.

```json
{
  "employee_name": {
    "value": "MICHAEL D BRYAN",
    "references": [
      "72ba3cca-01e5-407b-9fc4-81f54f9f0c51"
    ]
  },
  "employee_ssn": {
    "value": "555-50-1234",
    "references": [
      "a3f5d8c9-2b4e-4a1c-8f7e-9d6c5b4a3e2f"
    ]
  }
}
```

--------------------------------

### GET /v1/ade/extract/jobs/{job_id}

Source: https://docs.landing.ai/ade/ade-extract-troubleshoot

Troubleshooting guide for status codes returned by the Get Extract Job endpoint.

```APIDOC
## GET /v1/ade/extract/jobs/{job_id}

### Description
Retrieves the status and data of a specific extraction job.

### Method
GET

### Endpoint
/v1/ade/extract/jobs/{job_id}

### Parameters
#### Path Parameters
- **job_id** (string) - Required - The unique identifier of the extraction job.

### Response
#### Success Response (200)
- **status** (string) - The current status of the job (e.g., completed, failed).
- **data** (object) - The extracted data (included if status is completed).

#### Partial Content (206)
- **schema_violation_error** (string) - Details on schema non-conformance.
- **warnings** (array) - List of warnings regarding the extraction.

#### Error Responses
- **404 (Not Found)**: The job ID does not exist or belongs to a different API key.
- **500 (Internal Server Error)**: Server-side error during retrieval.
```

--------------------------------

### ADE Get Parse Jobs API

Source: https://docs.landing.ai/ade/ade-parse-troubleshoot

Troubleshooting guide for the ADE Get Parse Jobs API, including status code definitions and response field explanations.

```APIDOC
## GET /ade-get-parse-job

### Description
Retrieves the status and results of a previously submitted parse job.

### Method
GET

### Status Codes
- **200 (Success)** - Job status retrieved successfully. Check the `status` field.
- **206 (Partial Content)** - Job completed but some pages failed. Review `failed_pages` and `failure_reason`.
- **404 (Not Found)** - Job ID not found. Verify the ID and API key.
- **422 (Unprocessable Entity)** - Input validation failed. Verify the job ID format.
- **500 (Internal Server Error)** - Server error. Retry or contact support.

### Job Status Values
- `pending`: Job is queued.
- `processing`: Job is being processed.
- `completed`: Job finished successfully.
- `failed`: Job failed. Check `failure_reason`.
- `cancelled`: Job was cancelled.

### Response Fields
- **status** (string) - Current state of the job.
- **data** (object) - `ParseResponse` object for results < 1 MB.
- **output_url** (string) - Presigned S3 URL for results >= 1 MB.
- **progress** (number) - Completion percentage (0 to 1).
- **failure_reason** (string) - Details if the job failed.
- **failed_pages** (array) - List of pages that failed during processing.
```

--------------------------------

### Initialize Client with .env File

Source: https://docs.landing.ai/ade/agentic-api-key

Load the .env file before initializing the client to populate the environment variables.

```python
from dotenv import load_dotenv
from landingai_ade import LandingAIADE

load_dotenv()  # Populates the environment from .env.

client = LandingAIADE()
```

--------------------------------

### GET /v1/ade/extract/jobs

Source: https://docs.landing.ai/ade/ade-extract-troubleshoot

Troubleshooting guide for status codes returned by the List Extract Jobs endpoint.

```APIDOC
## GET /v1/ade/extract/jobs

### Description
Retrieves a list of extraction jobs.

### Method
GET

### Endpoint
/v1/ade/extract/jobs

### Parameters
#### Query Parameters
- **page** (integer) - Optional - The page number for pagination.
- **pageSize** (integer) - Optional - The number of items per page.
- **status** (string) - Optional - Filter jobs by status.

### Response
#### Success Response (200)
- **jobs** (array) - A list of extraction jobs.

#### Error Responses
- **422 (Unprocessable Entity)**: Invalid query parameters provided.
- **500 (Internal Server Error)**: Server-side error during listing.
```

--------------------------------

### Initialize Client and Execute Parsing

Source: https://docs.landing.ai/ade/ade-parse-visualize-sample

Sets up the LandingAI ADE client and triggers the parsing and visualization workflow for a specific file.

```typescript
// Initialize client (uses the API key from the VISION_AGENT_API_KEY environment variable)
  const client = new LandingAIADE();

  async function visualizeChunks() {
    // Replace with your file path
    const documentPath = "/path/to/file/document";

    // Parse the document
    console.log("Parsing document...");
    const parseResponse = await client.parse({
      document: fs.createReadStream(documentPath),
      model: "dpt-2-latest"
    });
    console.log("Parsing complete!");

    // Draw bounding boxes and create annotated images
    await drawBoundingBoxes(parseResponse, documentPath);
  }

  visualizeChunks();
```

--------------------------------

### Correct .env File Format

Source: https://docs.landing.ai/ade/agentic-api-key

Demonstrates the correct syntax for defining the API key in a .env file without spaces around the equals sign.

```text
# Correct
VISION_AGENT_API_KEY=<your-api-key>

# Incorrect
VISION_AGENT_API_KEY = <your-api-key>
```

--------------------------------

### Initialize ADE Client for EU

Source: https://docs.landing.ai/ade/ade-eu

Configure the client to use the EU deployment by setting the environment parameter during initialization.

```python
from landingai_ade import LandingAIADE

client = LandingAIADE(
    environment="eu",
)
```

```typescript
import LandingAIADE from "landingai-ade";

const client = new LandingAIADE({
  environment: "eu"
});
```

--------------------------------

### Configure Split Parameters in Python

Source: https://docs.landing.ai/ade/ade-python

Demonstrates basic usage of the split function with custom Split Rules.

```python
import json
from pathlib import Path
from landingai_ade import LandingAIADE

client = LandingAIADE()

split_response = client.split(
    split_class=json.dumps([
        {"name": "Section A", "description": "Introduction section"},
        {"name": "Section B", "description": "Main content section"}
    ]),
    markdown=Path("/path/to/parsed_output.md"),
    model="split-latest"
)
```

--------------------------------

### Initialize LandingAI ADE Client for EU

Source: https://docs.landing.ai/ade/ade-python

Set the environment parameter to 'eu' during client instantiation to route requests to European endpoints.

```python
from pathlib import Path
from landingai_ade import LandingAIADE

client = LandingAIADE(
    environment="eu",
)

# ... rest of your code
```

--------------------------------

### Initialize Client with Environment Variable

Source: https://docs.landing.ai/ade/agentic-api-key

The client automatically reads the VISION_AGENT_API_KEY environment variable upon initialization.

```python
from landingai_ade import LandingAIADE

client = LandingAIADE()
```

--------------------------------

### Parse Job Response

Source: https://docs.landing.ai/ade/ade-parse-async

Example JSON response containing the job ID.

```json
{'job_id': 'cmfx34ewm0000hyoqkh9dzd8n'}
```

--------------------------------

### Asynchronous Document Parsing with AsyncLandingAIADE

Source: https://docs.landing.ai/ade/ade-parse-async-sample

Initializes the async client and demonstrates parsing a document while saving the resulting markdown output.

```python
import asyncio
from pathlib import Path
from landingai_ade import AsyncLandingAIADE

# Initialize client (uses the API key from the VISION_AGENT_API_KEY environment variable)
client = AsyncLandingAIADE()

async def main() -> None:
    response = await client.parse(
        document=Path("path/to/file"),
        model="dpt-2-latest"
    )
    print(response.chunks)

    # Save markdown output (useful if you plan to run extract on the markdown)
    with open("output.md", "w", encoding="utf-8") as f:
        f.write(response.markdown)

asyncio.run(main())
```

--------------------------------

### Reload Claude Code Plugins

Source: https://docs.landing.ai/ade/build-with-ai-agents

Command to activate installed plugins in Claude Code.

```bash
/reload-plugins
```

--------------------------------

### GET /ade/parse/jobs

Source: https://docs.landing.ai/ade/ade-parse-troubleshoot

Retrieve a list of parse jobs with support for pagination and status filtering.

```APIDOC
## GET /ade/parse/jobs

### Description
Retrieve a list of parse jobs. Use this endpoint to monitor the status of your jobs.

### Parameters
#### Query Parameters
- **page** (integer) - Optional - Page number for pagination (default: 1)
- **page_size** (integer) - Optional - Number of jobs per page (default: 10)
- **status** (string) - Optional - Filter by job status (pending, processing, completed, failed, cancelled)

### Response
#### Success Response (200)
- **jobs** (array) - List of job summaries containing job_id, status, received_at, and progress
- **has_more** (boolean) - Indicates if more pages are available
- **failure_reason** (string) - Error details for failed jobs

### Status Codes
- **200** - Success: Job list retrieved successfully.
- **500** - Internal Server Error: Server error during job listing.
```

--------------------------------

### Initialize the ADE Client

Source: https://docs.landing.ai/ade/ade-typescript

Initialize the client using either automatic environment variable detection or explicit configuration.

```typescript
import LandingAIADE from "landingai-ade";

const client = new LandingAIADE();
```

```typescript
import LandingAIADE from "landingai-ade";

const client = new LandingAIADE({
  apikey: process.env.VISION_AGENT_API_KEY
});
```

--------------------------------

### Execute Get Results Script

Source: https://docs.landing.ai/ade/ade-parse-async

Command to run the Python script for retrieving parsing results.

```bash
python get-parse-results.py
```

--------------------------------

### Access ParseResponse Fields

Source: https://docs.landing.ai/ade/ade-python

Examples for filtering and accessing specific data within the ParseResponse object.

```python
for chunk in response.chunks:
    if chunk.type == 'text':
        print(f"Chunk {chunk.id}: {chunk.markdown}")
```

```python
page_0_chunks = [chunk for chunk in response.chunks if chunk.grounding.page == 0]
```

```python
for chunk in response.chunks:
    box = chunk.grounding.box
    print(f"Chunk at page {chunk.grounding.page}: ({box.left}, {box.top}, {box.right}, {box.bottom})")
```

```python
for chunk_id, grounding in response.grounding.items():
    print(f"Chunk {chunk_id} has type: {grounding.type}")
```

--------------------------------

### Configure Concise Image Descriptions

Source: https://docs.landing.ai/ade/ade-parse-custom-prompts

Limits the description length to 50 characters to reduce token usage.

```bash
custom_prompts={"figure":"Limit the description to 50 characters. Be concise but communicate the core message of the image. Do not truncate sentences."}
```

--------------------------------

### Schema Validation Error Response

Source: https://docs.landing.ai/ade/ade-extract-schema-json

Example of the error returned when the provided JSON schema is invalid.

```json
{
  "error": "Invalid JSON schema provided for fields_schema."
}
```

--------------------------------

### Define Extraction Schema

Source: https://docs.landing.ai/ade/ade-extract-response

Example of a JSON schema definition used to structure the extracted data.

```json
{
  "type": "object",
  "properties": {
    "employee_name": {
      "type": "string",
      "description": "The employee's full name"
    },
    "employee_ssn": {
      "type": "string",
      "description": "The employee's Social Security Number"
    },
    "gross_pay": {
      "type": "number",
      "description": "The gross pay amount"
    }
  }
}
```

--------------------------------

### Set API Key in Notebook

Source: https://docs.landing.ai/ade/agentic-api-key

Use the %env magic command to set the API key for the current notebook session.

```python
%env VISION_AGENT_API_KEY=<your-api-key>
```

```python
from landingai_ade import LandingAIADE

client = LandingAIADE()
```

--------------------------------

### Access Parse Response Data

Source: https://docs.landing.ai/ade/ade-typescript

Examples for iterating through chunks, filtering by page, and accessing grounding metadata.

```typescript
for (const chunk of response.chunks) {
  if (chunk.type === 'text') {
    console.log(`Chunk ${chunk.id}: ${chunk.markdown}`);
  }
}
```

```typescript
const page0Chunks = response.chunks.filter(chunk => chunk.grounding.page === 0);
console.log(page0Chunks);
```

```typescript
for (const chunk of response.chunks) {
  const box = chunk.grounding.box;
  console.log(`Chunk at page ${chunk.grounding.page}: (${box.left}, ${box.top}, ${box.right}, ${box.bottom})`);
}
```

```typescript
for (const [chunkId, grounding] of Object.entries(response.grounding)) {
  console.log(`Chunk ${chunkId} has type: ${grounding.type}`);
}
```

--------------------------------

### Initialize Client for EU Endpoints

Source: https://docs.landing.ai/ade/ade-typescript

Configure the client to use EU endpoints by setting the environment parameter during initialization.

```typescript
import LandingAIADE from "landingai-ade";

const client = new LandingAIADE({
  environment: "eu",
});

// ... rest of your code
```

--------------------------------

### Use Standardized Description String

Source: https://docs.landing.ai/ade/ade-parse-custom-prompts

Forces the model to return a fixed, static string for every figure.

```bash
custom_prompts={"figure":"Only return this string as the description: This is an image"}
```

--------------------------------

### Execution and Output Verification

Source: https://docs.landing.ai/ade/ade-quickstart

Commands to run the script and the expected output format.

```bash
python quickstart.py
```

```text
Extracted fields:
{'name': 'Sarah J. Mitchell', 'number_deposits': 5}
```

--------------------------------

### Chain Parse and Extract Operations

Source: https://docs.landing.ai/ade/ade-sf-extract-cloud

Example of calling api.parse and passing the resulting object directly into api.extract.

```sql
DECLARE
    parse_result OBJECT;
    extract_result OBJECT;
BEGIN
    -- Parse the document
    CALL api.parse(
        'https://ade.landing.ai/pdfs/invoice_1.pdf'
    ) INTO :parse_result;

    -- Extract using the parse result directly
    CALL api.extract(
        :parse_result,
        'https://ade.landing.ai/pdfs/InvoiceExtractionSchema.json'
    ) INTO :extract_result;

    RETURN extract_result;
END;
```

--------------------------------

### Parse with Optional Parameters

Source: https://docs.landing.ai/ade/ade-sf-parse-cloud

Configure the parsing process by specifying both the model version and a custom output table.

```sql
CALL api.parse(
    file_path => 'https://ade.landing.ai/pdfs/invoice_1.pdf',
    model => 'dpt-2-latest',
    output_table => 'invoice_results'
);

SELECT * FROM db.invoice_results;
```

--------------------------------

### Retrieve chunk confidence score

Source: https://docs.landing.ai/ade/ade-json-response

Example of the grounding object structure containing the confidence score for a chunk.

```json
"grounding": {
    "42ca60d3-a606-4c9a-a61e-493966b63fd9": {
      "box": {
        "left": 0.8607578277587891,
        "top": 0.9346558451652527,
        "right": 0.9316827058792114,
        "bottom": 0.951895534992218
      },
      "page": 0,
      "type": "chunkMarginalia",
      "confidence": 0.991,
      "low_confidence_spans": []
    }
}
```

--------------------------------

### Root Object Parsing Error Message

Source: https://docs.landing.ai/ade/ade-extract-troubleshoot

Example of the error message returned when the root of the JSON is not an object.

```text
Field extraction validation error: The provided JSON must parse to an object at the root.
```

--------------------------------

### Run the Grounding Script

Source: https://docs.landing.ai/ade/ade-extract-grounding-sample

Execute the grounding script from the current directory using either Python or TypeScript.

```bash
python grounding.py
```

```bash
npx tsx grounding.ts
```

--------------------------------

### Document Extraction JSON Response

Source: https://docs.landing.ai/ade/ade-quickstart

Example of the structured JSON output returned after performing document extraction.

```json
{
            "extraction":{
              "name":"Sarah J. Mitchell",
              "number_deposits":5
            },
            "extraction_metadata":{
              "name":{
                "references":[
                  "8b70276e-a3ee-4ed4-9f11-15439c50d603"
                ],
                "value":"Sarah J. Mitchell"
              },
              "number_deposits":{
                "references":[
                  "d9d59d6d-cf43-4ed2-8632-5f48d08885c5"
                ],
                "value":5
              }
            },
            "metadata":{
              "filename":"markdown-bank-statement.md",
              "org_id":"lgy5xucm9xnq",
              "duration_ms":6516,
              "credit_usage":1.5188,
              "job_id":"1bdcf9c99a614d698f3b4aca65f27935",
              "version":"extract-20260314",
              "schema_violation_error":null,
              "fallback_model_version":null,
              "warnings":[
                
              ]
            }
          }
```

--------------------------------

### Store API Key in .env File

Source: https://docs.landing.ai/ade/agentic-api-key

Use python-dotenv to load API keys from a local .env file into the environment.

```bash
pip install python-dotenv
```

```bash
VISION_AGENT_API_KEY=<your-api-key>
```

--------------------------------

### Set the Model via Python Library

Source: https://docs.landing.ai/ade/ade-parse-models

Pass the 'model' argument to the 'parse()' method of the LandingAIADE client. The library defaults to the latest 'dpt-2' snapshot if no model is specified.

```python
from pathlib import Path
from landingai_ade import LandingAIADE

client = LandingAIADE()

response = client.parse(
    document=Path("/path/to/document.pdf"),
    model="dpt-2-latest"
)
```

--------------------------------

### Extraction Output Format

Source: https://docs.landing.ai/ade/ade-parse-async

Example of the JSON structure returned by the API, containing extracted fields and metadata.

```json
{
  "extraction": {
    "exam_date": "2010-05-20",
    "procedure": "MRI OF THE LUMBAR SPINE WITH AND WITHOUT CONTRAST"
  },
  "extraction_metadata": {
    "exam_date": {
      "references": [
        "b5183837-035c-4a54-b324-a4c9e8a68027",
        "1-3"
      ],
      "value": "2010-05-20"
    },
    "procedure": {
      "references": [
        "56340fbe-8a51-46a0-a309-d91abd9b8b00"
      ],
      "value": "MRI OF THE LUMBAR SPINE WITH AND WITHOUT CONTRAST"
    }
  },
  "metadata": {
    "filename": "markdown-mri-report.md",
    "org_id": null,
    "duration_ms": 15080,
    "credit_usage": 1.3256,
    "job_id": "dfef14d3b66045b8abaf39788b9d17e8",
    "version": "extract-20260314",
    "schema_violation_error": null,
    "fallback_model_version": null,
    "warnings": []
  }
}
```

--------------------------------

### Nested Schema Metadata

Source: https://docs.landing.ai/ade/ade-extract-response

Example of metadata structure for nested extraction schemas, maintaining the original hierarchy.

```json
{
  "patient_details": {
    "patient_name": {
      "value": "John Smith",
      "references": [
        "72ba3cca-01e5-407b-9fc4-81f54f9f0c51"
      ]
    },
    "date": {
      "value": "2024-01-15",
      "references": [
        "72ba3cca-01e5-407b-9fc4-81f54f9f0c51"
      ]
    }
  },
  "emergency_contact_information": {
    "emergency_contact_name": {
      "value": "Jane Smith",
      "references": [
        "5b8865b9-1a81-46df-bcf7-0bdbed9130dc"
      ]
    },
    "relationship_to_patient": {
      "value": "Spouse",
      "references": [
        "5b8865b9-1a81-46df-bcf7-0bdbed9130dc"
      ]
    }
  }
}
```

--------------------------------

### Extracted Data Response

Source: https://docs.landing.ai/ade/ade-extract-response

Example of the structured data returned within the extraction field of the API response.

```json
{
  "employee_name": "MICHAEL D BRYAN",
  "employee_ssn": "555-50-1234",
  "gross_pay": 6000.00
}
```
Tokens:
Raw