Nutrient Data Extraction API

About Nutrient Data Extraction API

Nutrient Data Extraction API is a cloud API for parsing PDFs, scanned documents, images, and Office files into structured data. It converts documents into spatial JSON or Markdown, identifying elements such as tables, forms, formulas, images, charts, handwriting, key-value regions, headings, lists, and reading order, and returns each extracted value with bounding boxes, confidence scores, and page position.

The API supports mapping extracted data to a user-defined JSON Schema, with a schema generator available in the product's Studio interface to scaffold schemas from example documents. It offers multiple processing modes to balance speed, cost, and extraction depth, and handles multilingual documents.

It is intended for building document workflows that feed downstream systems such as databases, ERPs, CRMs, review queues, AI agents, retrieval-augmented generation pipelines, and search indexes. Output is designed to support validation and review through confidence scores, coordinates, and source evidence before being used in automated processes.

Key features

• Parses PDFs, scans, images, and Office files into spatial JSON or Markdown
• Detects tables, forms, formulas, images, charts, handwriting, key-value regions, headings, and lists
• Preserves page position, confidence scores, and reading order for extracted elements
• Maps extracted data to a custom JSON Schema, with a schema generator in Studio
• Extracts complex tables with rows, columns, spans, captions, and footnotes preserved
• Offers multiple processing modes to choose speed, cost, or depth

Frequently asked questions