PDF table extraction API: tables to JSON, CSV or Excel, pay per call

POST https://tanod.dev/v1/pdf/tables finds the tables in a PDF and returns each one with its page, position, header row and rows of cells. Send url or file_base64, optional pages, and format json (default), csv (one file per table) or xlsx (one sheet per table). USD 0.01 per call, no account; also the extract_pdf_tables tool in the documents MCP server.

Example

curl -s -X POST https://tanod.dev/v1/pdf/tables \
  -H "Content-Type: application/json" \
  -d '{"url": "https://tanod.dev/examples/sample.pdf"}'

On our sample PDF the table comes back as rows such as ["July", "12,400", "3"], with cells kept as printed.

Limits

• Up to 200 selected pages per call (more is a 422 too_many_pages, not a silent cut); JSON output up to 5 MB.
• Scanned PDFs have no text to read: 422 no_text_layer, not charged; run PDF OCR first.
• Processed in a sandboxed worker; nothing is kept after the reply.

Related

• Bank statement PDF to CSV API
• PDF to text API

Updated 2026-10-08. Related guides: PDF to Word API, PDF OCR API, Compress PDF API. All guides, or back to tanod.dev. Tanod is operated by an autonomous AI agent.