Skip to content

PDF Parser

Extract text from a PDF.

Reads a PDF from the input and returns its text content. It runs entirely in the browser.

Use this node to turn an uploaded or fetched PDF into text you can search, summarise, feed to an AI node, or store.

SettingNotes
Input fieldThe field holding the PDF data. Accepts base64, a data: URL, an ArrayBuffer/typed array, or a byte array. Defaults to data.
Output field nameThe field the extracted text is written to. Defaults to text.
Merge pagesWhen on (default), all pages are joined into one text string. When off, the output is an array with one string per page.

Returns { [outputFieldName]: text, pageCount }, where text is a single string (merge pages on) or an array of per-page strings (merge pages off), and pageCount is the number of pages.

  • No credential is required. Parsing runs in the browser.

Fetch or upload a PDF so its bytes land in a field, connect it into PDF Parser, then pass the extracted text to an AI Agent to summarise, or to Edit Fields for further processing.

  • If parsing fails, check that the input field actually contains the PDF (base64, a data: URL, or raw bytes) — an undecodable value throws an error.
  • Scanned PDFs that contain images rather than real text will return little or no text; use an OCR step upstream for those.