What is xParse?
xParse converts PDFs, images, and Office files into Markdown and structured JSON. Use Parse to preserve full document content and structure, including headings, tables, formulas, reading order, and element coordinates. Use Extract to return specific fields that match your extraction schema. Common workflows include RAG pipelines, document automation, and structured data extraction.Choose a workflow
Parse documents
Convert full documents into Markdown and structured JSON while preserving headings, tables, formulas, reading order, and element coordinates.Use Parse when you need:
- Full document content for RAG or search
- Structured document elements with coordinates
- Preserved tables, images, and formulas
Extract structured data
Extract only the fields you need and return structured JSON that matches your schema.Use Extract when you need:
- Specific fields from forms, invoices, or other business documents
- Structured data that matches your application schema
- Targeted output instead of full document content
What you can process
Supported input formats
Parse returns
- Markdown with preserved reading order
- Structured document elements
- Coordinates for each element
- Tables with preserved row and column structure, when available
- Structured output for images, formulas, handwriting, and form fields
Extract returns
- JSON matching your schema
- Source citations with page numbers and bounding boxes
- Optional confidence scores and source coordinates, when available
Common use cases
- Prepare documents for RAG pipelines
- Convert research papers into Markdown for search and analysis
- Extract fields from forms, invoices, receipts, and contracts
- Review parsed or extracted data against source documents
Get started
Try xParse in the web app
Upload a document and inspect the parsed or extracted output without writing code.
Call the API
Use the API to parse a document or extract structured fields from your application.
Next steps
Authentication
Get your App ID and Secret Code.
Supported Files & Limits
Review supported formats, file sizes, page limits, and concurrency limits.
Parse Quickstart
Parse your first document.
Extract Quickstart
Extract structured data with an extraction schema.