Nutrient hosts the DWS Data Extraction API at https://api.nutrient.io. Use this HTTP API to extract structured content and domain-specific data from documents.
Base URL
Use this base URL for all Data Extraction API endpoints:
https://api.nutrient.ioAll endpoints are relative to this base URL.
Authentication
Include your API key in the Authorization header with every request:
Authorization: Bearer pdf_live_...Get API keys from the DWS API keys page(opens in a new tab). Live keys start with pdf_live_ and work with every DWS integration enabled for your account by default. Create a scoped live key when an integration requires restricted access.
Test keys are available only after Support enables them for your account. Use them for health checks, continuous integration and continuous delivery (CI/CD) pipelines, and integration testing with additional limitations.
New accounts receive a global secret API key for DWS APIs. For production, create scoped secret API keys for the products your backend uses, such as DWS Data Extraction API and DWS Processor API. If /extraction/* returns 403, verify that the key is enabled for Data Extraction API.
Keep secret API keys on your backend. DWS Viewer API also supports publishable API keys for standalone Web SDK authorization. Those keys can be embedded in frontend code when scoped to your domains.
Available endpoints
The API provides the endpoints below for parsing documents, extracting schema-shaped data, and classifying documents.
| Endpoint | Description |
|---|---|
POST /extraction/parse | Extracts structured elements or Markdown from documents. Supports four processing modes: text, structure, understand, and agentic. Supports spatial elements and Markdown output. |
POST /extraction/extract | Extracts domain-specific JSON data from documents and maps it to your JSON Schema, with optional per-field citations. |
POST /extraction/classify | Scores a document against labels you define and returns a ranked list of predictions. It’s zero-shot, with no training data or templates. |
Further details
Use these guides to continue configuring the Data Extraction API:
- Refer to the supported languages guide for the full list of 100+ optical character recognition (OCR) languages with ISO codes and aliases.
- Refer to the supported file types guide for PDFs, images, and Office files accepted by the API.
- Refer to the error handling guide for HTTP status codes, error response formats, and troubleshooting.