Table Extract PDF MCP
Extract tables from PDF documents as structured JSON. Handles complex layouts including merged cells, multi-row headers, spanning columns, and nested tables. Returns normalized tabular data with position metadata.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
pdf_url |
string | ✅ | PDF URL or base64 encoded content |
pages |
array | ❌ | Specific pages to process (default: all) |
normalize_headers |
boolean | ❌ | Clean and standardize column headers (default: true) |
output_format |
string | ❌ | json, csv, or excel (default: json) |
Example Response
{
"total_tables": 7,
"tables": [
{
"table_id": 1,
"page": 3,
"caption": "Table 2: Financial Summary FY2023",
"headers": ["Metric", "FY2022", "FY2023", "YoY Change"],
"rows": [
["Revenue", "$84.2M", "$102.7M", "+21.9%"],
["EBITDA", "$12.1M", "$18.4M", "+52.1%"]
],
"merged_cells": 0,
"confidence": 0.97
}
]
}
Use Cases
- Financial statement normalization for M&A analysis
- Regulatory filing data extraction at scale
- Scientific paper data extraction for research databases
Access
# x402 endpoint — pay $0.05 USDC per call (Base mainnet)
POST https://x402.ntriq.co.kr/table-extract-pdf
# Service catalog
curl https://x402.ntriq.co.kr/services
x402 micropayments — USDC on Base, gasless EIP-3009