/jump to repo

Repos Recent Bookmarks Watched Notes Tags Discover Compare Stats GitHub

Repository brief

docling-project/docling

Read the upstream summary on the left, browse the cached forks below it, and load each fork comparison into the right-hand panel.

Cached analysis

cached 2026-03-30T11:43:48.674Z

3mo ago

docling-project/docling

Docling is a Python SDK and CLI for parsing documents into a unified representation for downstream gen AI workflows. The repository is active, production-stable, and very widely adopted, with 56,745 stars, 3,854 forks, and commits as recent as 2026-03-30. It supports many input formats, strong PDF understanding, OCR, export formats, local execution, and integrations with agentic AI tools.

Loading tags...

Stars56,745

Forks3,854

Default branchmain

Last pushed2026-03-30T09:09:25Z

Recommended shortcuts

Jump straight into Discofork's strongest cached fork picks, or open a compare view in one click.

Forks

Choose a fork to inspect

10 of 10 fork briefs

Maintenance:

Magnitude:

Sort:

Selected

Prefer this fork if your priority is a tailored parser with environment-specific path handling and internal workflow support. Prefer upstream if you want the newest Docling capabilities, lower maintenance burden, and easier upgrades.

Choose this fork only if its CSV and runtime-control additions are worth accepting an old, heavily diverged codebase. For most new adopters, upstream is the safer choice because it is far more current and actively maintained.

Prefer upstream unless you specifically need this exact older snapshot. This fork adds no visible functionality, and its main tradeoff is falling behind active upstream improvements.

Prefer this fork only if its local test corpus or workflow customizations are the asset you need. For most adopters, upstream is the better choice because this fork is stale, materially diverged, and likely missing recent bugfixes and capabilities.

Prefer upstream unless you specifically need the fork’s older CLI/PDF behavior or its lighter maintenance footprint. This fork looks materially behind and is a poor choice for users who want current Docling features, fixes, and integrations.

Choose this fork only if you want an older, debug-oriented Docling snapshot and can accept being materially behind upstream. If you want current format support, stability fixes, or ongoing maintenance, upstream is the better default.

Prefer upstream unless you specifically need this exact January 2026 snapshot. This fork adds no visible capabilities and is materially behind on current fixes and features, so it is a poor default choice for production adopters.

Choose upstream unless you specifically need the fork's older MLX/VLM, OCR-plugin, or DoclingParseV4 work. This fork is materially stale and carries substantial compatibility and maintenance risk.

Prefer this fork only if its custom OCR/artifact/image workflows are the goal. If you want the current Docling feature set, active maintenance, and upstream compatibility, upstream is the safer choice.

Choose this fork only if you specifically need its older behavior plus the small set of local fixes; otherwise upstream is the better default because this fork is stale and materially behind.

Fork comparison

genonai/doc_parser

70/100

active

significant_divergence

Prefer this fork if your priority is a tailored parser with environment-specific path handling and internal workflow support. Prefer upstream if you want the newest Docling capabilities, lower maintenance burden, and easier upgrades.

Likely purpose

To adapt Docling into a production parser for a specific GenAI stack, likely optimizing how documents and images are ingested in local, Docker, and Hugging Face-based deployments.

Best for

Teams that need a Docling-derived parser adapted to their own infrastructure and file layout.; Users running document parsing inside Docker or mixed local/Hugging Face environments.; Engineering teams willing to trade upstream freshness for internal workflow compatibility.

Additional features

More robust path resolution across deployment modes, including `huggingface_path`, `local_path(pathlib)`, and Docker-specific handling.
An image registration workflow with improved database query handling and error checks.
Dependency/version checkpointing for `unstructured` and `unstructured-inference`, suggesting controlled rollouts in a managed pipeline.
Project-specific packaging and branch workflow on `develop` for ongoing internal adaptation.
Introduces more operational workflow surface such as server-side handling, auth, session control, or admin tooling.
Changes agent workflow behavior, such as prompt handling, handoffs, worktrees, memory, or orchestration flow.

Missing features

It is 200 commits behind upstream, so it likely misses recent Docling fixes and features such as the HTML backend, remote KServe OCR support, and recent PDFium/table/LaTeX bug fixes.
Upstream's latest stability and integration work may lag here, making this fork riskier if you need the newest OCR, PDF, or export behavior without manual reconciliation.
It trails upstream by 200 commits, so some recent upstream features and fixes are likely not present yet.

Strengths

Strong customization for deployment-specific path handling.
Clear signs of active maintenance and frequent bug-fixing.
Likely better fit for internal/controlled document-processing pipelines than a plain upstream checkout.

Risks

Large divergence from upstream makes rebasing and future upgrades expensive.
Being behind upstream can leave users without recent bug fixes and backend improvements.
The custom workflows may be more opinionated and harder to adopt as a drop-in replacement.

docling-project/docling · Discofork