{
  "generatedAt": "2026-07-10T13:37:49.489Z",
  "runUrl": null,
  "results": [
    {
      "vendor": "docai",
      "dataset": "bankstatemently",
      "docType": "statement",
      "n": 5,
      "failures": 0,
      "latencyMsP50": 3737,
      "evaluated": 0,
      "normalizedOverall": null,
      "normalizedFields": {
        "date": null,
        "description": null,
        "amount": null,
        "balance": null
      }
    },
    {
      "vendor": "docai",
      "dataset": "cord",
      "docType": "receipt",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 2969,
      "items": {
        "precision": 0.3079019073569482,
        "recall": 0.450199203187251,
        "f1": 0.3656957928802589,
        "gtItems": 251,
        "predItems": 367,
        "correct": 113
      },
      "totalsAccuracy": 0.8215962441314554
    },
    {
      "vendor": "docai",
      "dataset": "fatura",
      "docType": "invoice",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 2884,
      "fieldAccuracy": 0.9309815950920245,
      "fieldsScored": 652,
      "anls": 0.9376019685835638,
      "perField": {
        "customerName": {
          "accuracy": 0.6451612903225806,
          "n": 62
        },
        "issueDate": {
          "accuracy": 1,
          "n": 98
        },
        "dueDate": {
          "accuracy": 1,
          "n": 58
        },
        "currency": {
          "accuracy": 1,
          "n": 92
        },
        "subtotal": {
          "accuracy": 1,
          "n": 68
        },
        "tax": {
          "accuracy": 0.9736842105263158,
          "n": 38
        },
        "total": {
          "accuracy": 0.9125,
          "n": 80
        },
        "vendorName": {
          "accuracy": 0.7794117647058824,
          "n": 68
        },
        "invoiceNumber": {
          "accuracy": 1,
          "n": 88
        }
      }
    },
    {
      "vendor": "docai",
      "dataset": "fintabnet",
      "docType": "tables",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 2622,
      "teds": 0.3649189984215049,
      "steds": 0.42323292930392137
    },
    {
      "vendor": "docai",
      "dataset": "sroie",
      "docType": "receipt",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 2328,
      "fieldAccuracy": 0.5275,
      "fieldsScored": 400,
      "anls": 0.6620101276692425,
      "perField": {
        "merchant": {
          "accuracy": 0.38,
          "n": 100
        },
        "date": {
          "accuracy": 0.61,
          "n": 100
        },
        "address": {
          "accuracy": 0.36,
          "n": 100
        },
        "total": {
          "accuracy": 0.76,
          "n": 100
        }
      }
    },
    {
      "vendor": "compound",
      "dataset": "bankstatemently",
      "docType": "statement",
      "n": 5,
      "failures": 0,
      "latencyMsP50": 36054,
      "evaluated": 0,
      "normalizedOverall": null,
      "normalizedFields": {
        "date": null,
        "description": null,
        "amount": null,
        "balance": null
      }
    },
    {
      "vendor": "compound",
      "dataset": "cord",
      "docType": "receipt",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 3900,
      "items": {
        "precision": 0.3890784982935154,
        "recall": 0.4541832669322709,
        "f1": 0.4191176470588235,
        "gtItems": 251,
        "predItems": 293,
        "correct": 114
      },
      "totalsAccuracy": 0.6713615023474179
    },
    {
      "vendor": "compound",
      "dataset": "fatura",
      "docType": "invoice",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 12848,
      "fieldAccuracy": 0.9938650306748467,
      "fieldsScored": 652,
      "anls": 0.9938650306748467,
      "perField": {
        "customerName": {
          "accuracy": 0.9838709677419355,
          "n": 62
        },
        "issueDate": {
          "accuracy": 1,
          "n": 98
        },
        "dueDate": {
          "accuracy": 1,
          "n": 58
        },
        "currency": {
          "accuracy": 0.9891304347826086,
          "n": 92
        },
        "subtotal": {
          "accuracy": 1,
          "n": 68
        },
        "tax": {
          "accuracy": 0.9473684210526315,
          "n": 38
        },
        "total": {
          "accuracy": 1,
          "n": 80
        },
        "vendorName": {
          "accuracy": 1,
          "n": 68
        },
        "invoiceNumber": {
          "accuracy": 1,
          "n": 88
        }
      }
    },
    {
      "vendor": "compound",
      "dataset": "fintabnet",
      "docType": "tables",
      "n": 100,
      "failures": 14,
      "latencyMsP50": 23448,
      "teds": 0.7912710446662822,
      "steds": 0.8190572886774939
    },
    {
      "vendor": "compound",
      "dataset": "sroie",
      "docType": "receipt",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 3130,
      "fieldAccuracy": 0.8875,
      "fieldsScored": 400,
      "anls": 0.9387130585797183,
      "perField": {
        "merchant": {
          "accuracy": 0.81,
          "n": 100
        },
        "date": {
          "accuracy": 0.97,
          "n": 100
        },
        "address": {
          "accuracy": 0.79,
          "n": 100
        },
        "total": {
          "accuracy": 0.98,
          "n": 100
        }
      }
    },
    {
      "vendor": "textract",
      "dataset": "cord",
      "docType": "receipt",
      "n": 100,
      "failures": 4,
      "latencyMsP50": 2526,
      "items": {
        "precision": 0.7379032258064516,
        "recall": 0.8061674008810573,
        "f1": 0.7705263157894737,
        "gtItems": 227,
        "predItems": 248,
        "correct": 183
      },
      "totalsAccuracy": 0.9545454545454546
    },
    {
      "vendor": "textract",
      "dataset": "fatura",
      "docType": "invoice",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 1792,
      "fieldAccuracy": 0.9217791411042945,
      "fieldsScored": 652,
      "anls": 0.9217791411042945,
      "perField": {
        "customerName": {
          "accuracy": 0.8548387096774194,
          "n": 62
        },
        "issueDate": {
          "accuracy": 1,
          "n": 98
        },
        "dueDate": {
          "accuracy": 1,
          "n": 58
        },
        "currency": {
          "accuracy": 0.9891304347826086,
          "n": 92
        },
        "subtotal": {
          "accuracy": 0.8970588235294118,
          "n": 68
        },
        "tax": {
          "accuracy": 0.9473684210526315,
          "n": 38
        },
        "total": {
          "accuracy": 0.875,
          "n": 80
        },
        "vendorName": {
          "accuracy": 0.6764705882352942,
          "n": 68
        },
        "invoiceNumber": {
          "accuracy": 1,
          "n": 88
        }
      }
    },
    {
      "vendor": "textract",
      "dataset": "fintabnet",
      "docType": "tables",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 2457,
      "teds": 0.8364195677742853,
      "steds": 0.8910186803972542
    },
    {
      "vendor": "textract",
      "dataset": "sroie",
      "docType": "receipt",
      "n": 100,
      "failures": 0,
      "latencyMsP50": 2074,
      "fieldAccuracy": 0.6875,
      "fieldsScored": 400,
      "anls": 0.8307924188885697,
      "perField": {
        "merchant": {
          "accuracy": 0.74,
          "n": 100
        },
        "date": {
          "accuracy": 0.97,
          "n": 100
        },
        "address": {
          "accuracy": 0.08,
          "n": 100
        },
        "total": {
          "accuracy": 0.96,
          "n": 100
        }
      }
    }
  ]
}