1
Upload
PDF, Word (.docx), PowerPoint (.pptx), ODT, RTF, CSV, JSON, Markdown, or text.
Scanned PDFs are sent to the model as page images.
PDF to API
Turn PDFs into a queryable API, linked to the data they belong to.
Upload invoices, statements, or contracts. DocQuery's AI extraction turns each file into rows with typed columns, then serves them over REST, GraphQL, and MCP. Link the extracted rows to your customer list or contract tracker and every call can return both.
From file to endpoint
2
Review the tables
Extraction proposes tables and columns. Edit cells or re-run extraction
before you rely on the data.
3
Link
Connect a column to another table, for example invoices.customer_id → customers.id,
where customers comes from Excel, Airtable, or another file.
4
Query
Every table gets REST endpoints with column filters and
include=
for related rows, a GraphQL schema, an OpenAPI spec, and a hosted MCP server.
What a call looks like
Invoice rows extracted from PDFs, filtered on a column, with the linked customer embedded.
curl -H "Authorization: Bearer YOUR_API_KEY" \
"https://docquery.io/api/v1/tables/invoices/rows?status=unpaid&include=customer_id"Try a real call now, no signup. This uses the public read-only demo key (example customers and invoices):
curl -H "Authorization: Bearer dq_demo_readonly" \
"https://docquery.io/api/v1/tables/example_invoices/rows?status=due&include=customer_id"For AI agents, see the MCP server setup. For Airtable bases, see Airtable + PDFs.
When this is the right tool
- A few dozen to a few thousand documents that belong to records you already track.
- You want developers or an AI agent to query them, not just export them.
- If you need a high-volume accounts-payable inbox with approval workflows, a dedicated AP parsing tool will fit better.
Free plan: 5 documents and 10,000 rows, no card. Click Load example data on the dashboard to try linked tables right away.