Form Parser extracts key-value pairs (KVP), tables, selection marks (checkboxes),
and generic fields to augment and automate extraction. It can extract up to 11
generic entities and checkboxes out of the box. You don't specify the fields (schema),
you want to extract with the Form Parser. The model detects and returns entities
of interest from each page of documents.
Custom extractor
The custom extractor extracts entities you define in schema and offers three modeling options:
foundation model, custom model based, and custom template based. Given promising
results from foundation models with little to no training data, we recommend starting
with the foundation model as the first option and try out other options as needed.
The foundation models do zero- to few-shot prediction, based on up to 5 labeled
documents in the dataset, and fine-tuned prediction with more than 10 labeled documents in the dataset.
Training method
Document examples
Document layout variation
Free form text or paragraphs
Number of training documents for production-ready quality, depending on variability
Fine tune and foundation model (generative AI).
Contract, terms of service, invoice, bank statement, bill of lading, payslips.
High to Low (preferred).
High.
Medium: 0-50+ documents.
Custom model.
Model.
Similar forms with layout variation across years or vendors (for example, W9).
Low to medium.
Low.
High: 10-100+ documents.
Template.
Tax forms with a fixed layout (for example, Forms 941 and 709).
None.
Low.
Low (3 documents).
Because foundation models typically require fewer training documents, they're
recommended as the first option for all variable layouts.
Layout Parser
Layout Parser transforms documents in various formats into structured
representations, making content like paragraphs, tables, lists, and structural
elements like headings, page headers, and footers accessible, and creating
context-aware chunks that facilitate information retrieval in a range of
generative AI and discovery apps.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-08-26 UTC."],[[["\u003cp\u003eDocument AI offers tools like Form Parser, Custom extractor, and Layout Parser for extracting information from documents based on various use cases.\u003c/p\u003e\n"],["\u003cp\u003eForm Parser automatically extracts key-value pairs, tables, selection marks, and up to 11 generic entities without needing a predefined schema.\u003c/p\u003e\n"],["\u003cp\u003eThe Custom extractor allows users to define their extraction schema and offers three modeling options: foundation model, custom model-based, and custom template-based.\u003c/p\u003e\n"],["\u003cp\u003eFoundation models in Custom extractors are recommended as the first option due to their ability to perform with minimal training data.\u003c/p\u003e\n"],["\u003cp\u003eLayout Parser transforms documents into structured data, identifying elements such as paragraphs, tables, lists, headings, and headers/footers, for use in information retrieval and generative AI applications.\u003c/p\u003e\n"]]],[],null,["# Extraction overview\n===================\n\nDocument AI offers multiple products to extract information from documents\nfor different use cases:\n\n- [Form Parser](#form-parser)\n- Custom extractor, which offers three different modeling types:\n\n - Foundation model\n - Custom model based\n - Custom template based\n- [Layout Parser](#layout-parser)\n\nForm Parser\n-----------\n\nForm Parser extracts key-value pairs (KVP), tables, selection marks (checkboxes),\nand generic fields to augment and automate extraction. It can extract up to 11\ngeneric entities and checkboxes out of the box. You don't specify the fields (schema),\nyou want to extract with the Form Parser. The model detects and returns entities\nof interest from each page of documents.\n\nCustom extractor\n----------------\n\nThe custom extractor extracts entities you define in schema and offers three modeling options:\nfoundation model, custom model based, and custom template based. Given promising\nresults from foundation models with little to no training data, we recommend starting\nwith the foundation model as the first option and try out other options as needed.\nThe foundation models do zero- to few-shot prediction, based on up to 5 labeled\ndocuments in the dataset, and fine-tuned prediction with more than 10 labeled documents in the dataset.\n\nBecause foundation models typically require fewer training documents, they're\nrecommended as the first option for all variable layouts.\n\nLayout Parser\n-------------\n\n| **Note:** Layout Parser is in Public preview\n\nLayout Parser transforms documents in various formats into structured\nrepresentations, making content like paragraphs, tables, lists, and structural\nelements like headings, page headers, and footers accessible, and creating\ncontext-aware chunks that facilitate information retrieval in a range of\ngenerative AI and discovery apps."]]
