Parse Document
AI-generatedSummary
Extracts the full structured JSON from a receipt or invoice (image or PDF) by sending it to the ParseDoc API, and returns the parsed document as the item's JSON.
Inputs
- Document Source — Where the document comes from: 'Binary Property' (file from a previous node, the default) or 'URL' (public document URL).
- Binary Property (required) — Shown when source is Binary. Name of the binary property holding the image or PDF (default 'data'). The file is base64-encoded and sent with its MIME type, falling back to image/jpeg.
- Document URL (required) — Shown when source is URL. Public URL of the document, sent to the API as image_url.
Output shape
One output item per input item, whose JSON is the full 'document' object returned by ParseDoc. The source shows fields such as date, merchant, document_type, invoice_number, currency, subtotal, tax, total, category, payment_method, line_items and confidence.
Each output item is paired to its input item. If the API response is not ok or has no document, the node throws an error. With Continue On Fail enabled, it outputs {error: message} for that item instead. Without a ParseDoc API credential, the node calls the public /api/demo endpoint. With an API key, it calls /api/key/v1/parse using Bearer auth and the credential's base URL. Each request times out after 90 seconds.
Examples
Example 1: Parse a receipt downloaded earlier in the workflow
Document Source: Binary Property; Binary Property: data. Place it after a Read/Download node and use the returned line_items and totals downstream.
Example 2: Parse an invoice hosted at a public link
Document Source: URL; Document URL: https://example.com/invoice.pdf, or an expression pulling the URL from a previous item.