Back to prompts

Table Data Extractor

extraction0 saves

The Table Data Extractor converts textual descriptions of data into a structured table format, enabling users to easily organize and analyze information presented in prose, lists, or semi-structured formats. It infers column names when they are not explicitly provided, streamlining data extraction for various applications.

Prompt Text

Extract tabular data from the following text or description. The text may describe data in prose form, lists, or semi-structured format.

Input:
{{text}}

Convert this into a structured table format:
{
  "headers": ["column1", "column2", ...],
  "rows": [
    ["value1", "value2", ...],
    ...
  ]
}

Infer appropriate column names if not explicitly stated.

Evaluation Results

1/28/2026
Overall Score
4.52/5

Average across all 3 models

Best Performing Model
Low Confidence
openai:gpt-5-mini
5.00/5
openai:gpt-5-mini
#1 Ranked
5.00
/5.00
adh
1.9
cla
1.8
com
1.9
In
900
Out
2,850
Cost
$0.0059
google:gemini-2.5-flash-lite
#2 Ranked
4.44
/5.00
adh
1.9
cla
1.9
com
1.7
In
1,023
Out
972
Cost
$0.0005
anthropic:claude-3-5-haiku
#3 Ranked
4.11
/5.00
adh
1.8
cla
1.9
com
1.7
In
1,030
Out
958
Cost
$0.0047
Test Case:

Tags