compare
Docling vs Unstructured
The same facts for both, read from GitHub every night, and the relation a person reviewed.
| Fact | Docling | Unstructured |
|---|---|---|
| Language | Python | HTML |
| Licence | MIT | Apache-2.0 |
| Stars | 68k | 16k |
| Latest | v2.134.0 | 0.27.16 |
| Last push | 2026-10-06 | 2026-10-06 |
| Release cadence | about 2 days between releases | about 6 days between releases |
| Active contributors | 120 commit authors on the default branch in the last 90 days | 17 commit authors on the default branch in the last 90 days |
| Flags | none | none |
How they relate
Both replace LlamaParse. Alternatives to LlamaParse →
Full replacementDoclingParses PDF, DOCX, PPTX, HTML and images into Markdown or JSON with layout and table models.
PartialUnstructuredPartitions documents into typed elements for RAG pipelines, from Python.
Docling
- v2.134.02026-10-06iwork: Add Apple Numbers (.numbers) support (#4113) (8d7cc9f)
- v2.133.02026-10-03ocr: Allow configuring RapidOCR model size (#4500) (a1c5ff2)
- v2.132.02026-10-01pdf, reading-order: Updating the reading-order algoirthm (#4276) (42de1e2)
- v2.131.02026-09-29iwork: Parse Keynote charts as classified pictures with data (#4376) (9da6935)
- v2.130.02026-09-22Add Mineru 2.5 Pro to the supported VLMs (#4322) (9a2c3e7)
Unstructured
- 0.27.162026-10-05This release rolls up all changes since 0.27.10; versions 0.27.11–0.27.15 were not published to PyPI individually.
- 0.27.102026-09-27fix(html): extract definition lists instead of discarding them
- 0.27.92026-09-26fix(chunking): retain table rowspans across cell splits
- 0.27.82026-09-22fix: partition multi-section DOCX files in linear time
- 0.27.62026-09-14Fix DOCX text_as_html duplicating merged-cell text instead of colspan/rowspan
Docling
unsigned The latest release, v2.134.0, carries no signature GitHub could verify.
Loading the security report
Unstructured
✓ signed The latest release, 0.27.16, carries a signature GitHub verified.
Loading the security report