Back to Gap Map Model components / Document conversion & OCR

OpenDataLoader PDF

Hancom
open core / Overall score: 3.4

OpenDataLoader PDF converts PDFs into Markdown, HTML and JSON with a bounding box and type for every element, using a deterministic Java engine for reading order, headings, lists and tables. A hybrid mode sends complex pages to a local Docling-based backend for borderless tables, OCR, formulas and chart descriptions, and the tool can also tag untagged PDFs for accessibility. Hancom develops it, with Python, Node.js and Java packages.

Openness

4 high confidence
4.0
license
Apache-2.0 (OSI. the LICENSE body is the Apache License 2.0, and NOTICE names Hancom, Inc. as copyright holder)
source
public (opendataloader-project/opendataloader-pdf)
core features withheld
yes — the README's capability matrix marks PDF/UA-1 and PDF/UA-2 export and an accessibility studio as Enterprise, outside the Apache-2.0 source

The parser is released under the Apache License 2.0 with its source public, and every extraction feature is in the open code. PDF/UA export and the visual accessibility editor are sold as an enterprise add-on, so part of the product sits outside the published source.

Adoption

3 high confidence
3.0

Adoption is measured as PyPI downloads of opendataloader-pdf, the Python wrapper that bundles the Java engine and links back to the repository. The npm package is declared as well, and its downloads are not part of this figure.

Capability

4 high confidence
4.0

OpenDataLoader PDF keeps reading order, headings, lists and tables, and returns each element with its position and type. It sits a step below Docling because it returns no fields a user defines, such as an invoice total, and it reads PDFs only.

Verified 2026-11-12