The Retab API key is required (https://docs.retab.com/overview/quickstart)
Tool that utilizes Retab to extract information from documents.
Arguments:
json_schema
(string
, required): JSON schema defining the expected data structure.model
(string
, required): The model to use for extraction.document
(file
, required): The document to be extracted.modality
(select
, required): The modality of the document. Options:native
,image
,text
,image+text
.image_resolution_dpi
(number
, optional): The resolution the image is scanned at.browser_canvas
(select
, optional): The canvas size of the browser. Options:A3
,A4
,A5
.temperature
(number
, optional): The temperature of the model.n_consensus
(number
, optional): The number of consensus extractions to perform.reasoning_effort
(select
, optional): The effort level for the model to reason about the input data. Options:low
,medium
,high
.
Tool that utilizes Retab to parse documents into an LLM-friendly format.
Arguments:
model
(string
, required): The model to use for extraction.document
(file
, required): The document to be extracted.image_resolution_dpi
(number
, optional): The resolution the image is scanned at.browser_canvas
(select
, optional): The canvas size of the browser. Options:A3
,A4
,A5
.table_parsing_format
(select
, optional): The format to parse tables into. Options:markdown
,yaml
,html
,json
.
Github repository: https://github.com/retab-dev/dify Contact: sacha@retab.com