Skip to main content

Get Extraction Data

Retrieve the structured row data from an extraction, including per-cell provenance, confidence scores, and reasoning traces for every extracted value.

GET/v1/extractions/:id/data

Parameters

ParameterTypeRequiredDescription
idstring (path)YesThe extraction ID
include_provenanceboolean (query)NoInclude provenance objects (default: true)
formatstring (query)NoResponse format: json (default) or csv

Example

curl "https://api.talonic.com/v1/extractions/ext_abc123/data?include_provenance=true" \
  -H "Authorization: Bearer tlnc_sk_live_7f3a...x9k2"

Response

{
  "extraction_id": "ext_abc123",
  "schema_id": "sch_abc123",
  "rows": [
    {
      "cells": [
        {
          "field": "vendor_name",
          "value": "Acme Corp",
          "confidence": 0.98,
          "flagged": false,
          "provenance": {
            "page": 1,
            "bbox": [50, 120, 200, 140],
            "raw_text": "Acme Corporation GmbH",
            "reasoning": "Matched vendor label in header region"
          }
        },
        {
          "field": "total_amount",
          "value": 1250.00,
          "confidence": 0.97,
          "flagged": false,
          "provenance": {
            "page": 1,
            "bbox": [320, 450, 420, 470],
            "raw_text": "Total: EUR 1,250.00",
            "reasoning": "Matched 'Total' label adjacent to currency value"
          }
        }
      ]
    }
  ]
}

Understanding provenance

Each cell's provenance links to the exact source region in the original document. The bbox array contains [x1, y1, x2, y2] coordinates. The confidence gate (see pipeline) determines the flagged status. Read more about per-cell provenance, the Field Registry, and correcting values.