Document parsing for generative AI
Many formats in.
One document out.
PDFs, office files, images, audio and email arrive in a single expressive representation, which is what makes the next step in a pipeline tractable.
Formats it parses
- DOCX
- PPTX
- XLSX
- HTML
- EPUB
- images
- audio
- LaTeX
- and more
What you get
DoclingDocumentLayout, reading order, table structure, code and formulas, held in one model.
- Markdown
- HTML
- JSON
- DocTags
Formats listed are drawn from the project's own supported-formats list, not a benchmark.
Advanced PDF understanding covers page layout, reading order, table structure, code, formulas and image classification.