PDF
PDF Extract Text
Extracts the text of each upstream PDF and returns one row per page.
When to use it
Use it when you want to search, filter or parse the wording of PDFs rather than their tables, for example finding every document that mentions a clause.
In a pipeline
PDF File InputPDF Extract TextExport CSV
Example
One row per page of a contract
| File | Pages |
|---|---|
| contract.pdf | 3 |
| Page | Text | Character Count |
|---|---|---|
| 1 | SERVICE AGREEMENT — made on 1 August 2026 | 1842 |
| 2 | 3. Payment terms — invoices fall due 30 days from issue | 2110 |
| 3 | Signed for and on behalf of Northwind Ltd | 640 |
The node has no settings of its own; set the Page Range on the PDF File Input node upstream to limit which pages arrive.
Worth knowing
- Text is rebuilt from positioned items and grouped into lines, so a two-column page can interleave its columns.
- Rows are deduplicated by source file, so several page rows of one PDF still extract that PDF once.
- Each row carries __source_file and __pdf_path, which keeps later PDF nodes pointed at the same file.
- A scanned PDF with no text layer returns empty text; there is no OCR step in this node.