Skip to main content

extract_text

Extract the full text of a document using AI-powered OCR.

Unlike the specific parsers (parse_invoice, parse_nfe, ...), this returns the raw text of any document with no predefined schema — useful for feeding a document's contents to prompt for extraction or analysis.

Raises:

ValueError: If the file type is unsupported.

Parameters

NameDescriptionTypeDefault
document_path (required)The path to the document to be parsed. Supported types: .pdf, .jpeg, .jpg, .png.Union[Path, str]None

Return Value

str