NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI, #2152 most downloaded on PyPI
SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications.
Latest stable
2.150.1
Last release 7 days ago
14 Nov 2026
Ships on a steady schedule
a new release about every 9 days
Nearly every release is documented
notes for 60 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
2 years old
241 releases · first in 2024
Drag across the ruler to compare, or tap a release to open it.
Added support for exporting DocItem to an image when page image is available (#379) (`3f91e7d`)
3f91e7d)ed785ea)926dfd2)7a97d71)8533039)8b437ad)Your coding agent can read these notes before it upgrades. Set up the MCP server →
Skip glm model downloads (#322) (`c9341bf`)
OCR: Introduce the OcrOptions.force_full_page_ocr parameter that forces a full page OCR scanning (#290) (`c6b3763`)
c6b3763)5d4a10b)81c8243)97f214e)EasyOcrModel: Support the use_gpu pipeline parameter in EasyOcrModel. Initialize easyocr (#282) (`0eb065e`)
tesserocr: Raise Exception if tesserocr has not loaded any languages (#279) (`704d792`)
Pdf backend, table mode as options and artifacts path (#203) (`40ad987`)
Simplify torch dependencies and update pinned docling deps (#190) (`eb679cc`)
Add pipeline timings and toggle visualization, establish debug settings (#183) (`2a2c65b`)
Fix header levels for DOCX & HTML (#184) (`b9f5c74`)
b9f5c74)94d0729)7d19418)88c1673)Add coverage_threshold to skip OCR for small images (#161) (`b346faf`)
New experimental docling-parse v2 backend (#131) (`5e4944f`)
Remove stderr from tesseract cli and introduce fuzziness in the text validation of OCR tests (#138) (`dae2a3b`)
Add options for choosing OCR engines (#118) (`f96ea86`)
New torch-based docling models (#120) (`2422f70`)
Windows support (#122) (`d44c62d`)
Allow usage of opencv 4.6.x (#110) (`34bd887`)
Support tableformer model choice (#90) (`d6df76f`)
Add figure in markdown (#98) (`6a03c20`)
Updated the render_as_doctags with the new arguments from docling-core (#93) (`4794ce4`)
tests: Adjust the test data to match the new version of LayoutPredictor (#82) (`fa9699f`)
CLI compatibility with python 3.10 and 3.11 (#79) (`2870fdc`)
Add docling cli (#75) (`9899078`)
Adding txt and doctags output (#68) (`bdfdfbf`)
Linux arm64 support and reducing dependencies (#69) (`27a7a15`)
Add unit tests (#51) (`48f4d1b`)
Propagate row_section in tables (#57) (`de85e46`)
Table cells overlap and model warnings (#53) (`f49ee82`)
Refine conversion result (#52) (`e46a66a`)
Align output formats (#49) (`8cc147b`)
Page-level error reporting from PDF backend, introduce PARTIAL_SUCCESS status (#47) (`a294b7e`)
Better raise exception when a page fails to parse (#46) (`8808463`)
Upgrade docling-parse PDF backend and interface to use page-by-page parsing (#44) (`a8c6b29`)
Usage of bytesio with docling-parse (#43) (`fac5745`)
Remove [ocr] extra to fix wheel install (#42) (`6995268`)
Add scipy as dependency (#40) (`f19871a`)
Add adaptive OCR, factor out treatment of OCR areas and cell filtering (#38) (`e94d317`)
Allow computing page images on-demand with scale and cache them (#36) (`78347bf`)
Update parser with bytesio interface and set as new default backend (#32) (`90dd676`)
Output page images and extracted bbox (#31) (`63d80ed`)
Introducing docling_backend (#26) (`b8f5e38`)
Set page number using 1-based indexing (#22) (`d2d9543`)
Correct text extraction for table cells (#21) (`f4bf3d2`)
Add simplified single-doc conversion (#20) (`d603137`)
Add easyocr to main deps for valid extra (#19) (`54b3dda`)
Expose ocr as extra (#18) (`b0725e0`)
V1.0.0 release (#16) (`71c3a9c`)
Optimize table extraction quality, add configuration options (#11) (`e9526bb`)
Missing type for default values (#12) (`d1d1724`)
Enable python 3.12 support by updating glm (#8) (`fb72688`)
### Feature * Build with ci (#6) (`b1479cf`)
Nothing published for this version
First public release of docling
doclingOne column per month.