---
updatedAt: 2026-10-01T17:11:41.000Z
agentTools:
projectIndex: https://docs.sensible.so/llms.txt
---
# Verbosity
Configure the verbosity parameter for document extraction, including output levels that add source text lines, position data, and bounding polygon metadata.
Configures the verbosity of the extraction.
# Parameters
| Possible Values | Description |
| :-------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `0` | default |
| `1` | Adds a `lines` array to each field's output describing the field's source text in the document. Each line includes information about its position in the document, for example, `page`, and `boundingPoly` metadata. |
| `3` | - In addition to the `lines` array:
- Sensible returns `points` and `regions` arrays as siblings to the `lines` array
- If Sensible extracted a field from OCR'd text, then Sensible adds the following confidence scores to the output:
1. In the field's standardized text metadata, Sensible provides an OCR `confidence` score for each element in the `lines` array. The score is the average score of the "tokens", or words, in the line.
2. In the field output, Sensible outputs OCR `anchorConfidence` and `valueConfidence` values, which are the averages of their source lines' confidence scores.
Sensible's OCR engines output confidence scores between 0 and 1. For more information about using OCR confidence scores, see [Validate extractions](https://docs.sensible.so/docs/validate-extractions).
If the verbose output causes the extraction response to exceed 6MB, as can be the case with large documents, then Sensible returns a `413` error. |
# Examples
If you add `"verbosity": 1,` as a sibling to the fields array in the [Fixed Table](https://docs.sensible.so/docs/fixed-table#examples) example, you get the following output:
```json
{
"agile_risks_table": {
"columns": [
{
"id": "col1_risk_description",
"values": [
{
"value": "Poor task point estimation",
"type": "string",
"lines": [
{
"text": "Poor task point estimation",
"page": 0,
"boundingPolygon": [
{
"x": 0.993,
"y": 2.75
},
{
"x": 2.631,
"y": 2.753
},
{
"x": 2.63,
"y": 3.378
},
{
"x": 0.993,
"y": 3.376
}
]
}
]
},
{
"value": "Poor epic scope definition",
"type": "string",
"lines": [
{
"text": "Poor epic scope definition",
"page": 0,
"boundingPolygon": [
{
"x": 0.993,
"y": 3.376
},
{
"x": 2.63,
"y": 3.378
},
{
"x": 2.63,
"y": 4.004
},
{
"x": 0.992,
"y": 4.001
}
]
}
]
},
{
"value": "Inadequate scrum master training",
"type": "string",
"lines": [
{
"text": "Inadequate scrum master training",
"page": 0,
"boundingPolygon": [
{
"x": 0.992,
"y": 4.001
},
{
"x": 2.63,
"y": 4.004
},
{
"x": 2.63,
"y": 4.868
},
{
"x": 0.992,
"y": 4.866
}
]
}
]
}
]
},
{
"id": "col4_rank_last_month",
"values": [
{
"source": "2",
"value": 2,
"type": "number",
"lines": [
{
"text": "2",
"page": 0,
"boundingPolygon": [
{
"x": 5.878,
"y": 2.758
},
{
"x": 7.512,
"y": 2.761
},
{
"x": 7.511,
"y": 3.386
},
{
"x": 5.878,
"y": 3.383
}
]
}
]
},
{
"source": "1",
"value": 1,
"type": "number",
"lines": [
{
"text": "1",
"page": 0,
"boundingPolygon": [
{
"x": 5.878,
"y": 3.383
},
{
"x": 7.511,
"y": 3.386
},
{
"x": 7.511,
"y": 4.011
},
{
"x": 5.877,
"y": 4.008
}
]
}
]
},
{
"source": "3",
"value": 3,
"type": "number",
"lines": [
{
"text": "3",
"page": 0,
"boundingPolygon": [
{
"x": 5.877,
"y": 4.008
},
{
"x": 7.511,
"y": 4.011
},
{
"x": 7.511,
"y": 4.874
},
{
"x": 5.877,
"y": 4.872
}
]
}
]
}
]
}
],
"title": {
"type": "string",
"value": "Agile software development risk tracking",
"lines": [
{
"text": "Agile software development risk tracking",
"page": 0,
"boundingPolygon": [
{
"x": 0.994,
"y": 1.666
},
{
"x": 4.902,
"y": 1.673
},
{
"x": 4.902,
"y": 1.886
},
{
"x": 0.994,
"y": 1.879
}
]
}
]
},
"footer": {
"type": "string",
"value": "Project managers and technical program managers can expand on the preceding table for risk management in Agile software development.",
"lines": [
{
"text": "Project managers and technical program managers can expand on the preceding table for risk management in Agile software development.",
"page": 0,
"boundingPolygon": [
{
"x": 0.981,
"y": 5.161
},
{
"x": 7.447,
"y": 5.169
},
{
"x": 7.447,
"y": 5.663
},
{
"x": 0.981,
"y": 5.655
}
]
}
]
}
}
}
```