Veryfi icon

Veryfi

Extract data from invoices, receipts, checks, bank statements, W-9/W-2, IDs, and any document via Veryfi

Actions10

Document → Extract Data From Receipt

AI-generated

Summary

Sends a receipt image or PDF to Veryfi to extract structured data including line items, totals, and metadata.

Inputs

  • fileSource (required) — The source of the file: binary data from a previous node or a public URL.
  • binaryPropertyName (required) — The name of the binary field containing the receipt file (e.g., 'data') when using Binary File source.
  • fileUrl (required) — A public URL to the receipt file. Veryfi fetches it server-side when using URL source.
  • fileName — Optional custom name for the file sent to Veryfi. Defaults to the detected name from the binary input.
  • confidenceDetails — If true, Veryfi returns per-field confidence scores as objects containing value, score, and ocr_score.
  • flattenConfidence — When confidenceDetails is true, flattens the score objects into sibling keys (e.g., field__score). Defaults to true.
  • additionalOptions — A collection of optional settings: auto_categorize (default true), boost_mode, categories, external_id, and tags.

Output shape

A single item with extracted receipt data in JSON format. If confidenceDetails is enabled, fields contain nested objects.

When confidenceDetails is true, the node uses the transformConfidence helper to flatten score objects into sibling keys (e.g., total__score) for easier mapping. Boost mode returns data without storing the document image.

Examples

Example 1: Extract data from a binary file uploaded via HTTP Request node

File Source: Binary File, Binary Property: data, Confidence Details: false

Example 2: Extract data from a public URL with confidence scores

File Source: URL, Confidence Details: true, Flatten Confidence Scores: true

Example 3: Extract data with specific category constraints

File Source: Binary File, Additional Options: Categories: ['receipt', 'invoice']

Links

Discussion