Unstructured-IO /unstructured
Convert documents to structured data effortlessly. Unstructured is open-source ETL solution for transforming complex documents into clean, structured formats for language models. Visit our website to learn more about our enterprise grade Platform product for production grade workflows, partitioning, enrichments, chunking and embedding.
★ 15,743⑂ 1,399👁 75⑃ 3,321● 1,299▤ 256
+3
Daily
+67
Weekly
+184
Monthly
15,743
stars
1,399
forks
75
watchers
3,321
prs
1,299
issues
256
releases
README
!https://pypi.python.org/pypi/unstructured/
!https://pypi.python.org/pypi/unstructured/
!https://GitHub.com/unstructured-io/unstructured.js/graphs/contributors
!code_of_conduct.md
!https://GitHub.com/unstructured-io/unstructured.js/releases
!https://github.com/Naereen/badges/
Open-Source Pre-Processing Tools for Unstructured Data
The unstructured library provides open-source components for ingesting and pre-processing images and text documents, such as PDFs, HTML, Word docs, and many more. The use cases of unstructured revolve around streamlining and optimizing the data processing workflow for LLMs. unstructured modular functions and connectors form a cohesive system that simplifies data ingestion and pre-processing, making it adaptable to different platforms and efficient in transforming unstructured data into structured outputs.
[!NOTE]
The open-source library is, and will stay, completely free. If you need higher-quality output, though, our latest models are available through the Transform v2 API. Compared with this library, Transform gets 2x table content accuracy, 58% less invented content, and 0.878 text accuracy (vs. 0.715):
| Model | Table Cell Content | Table Cell Spatial | Adjusted CCT (text accuracy) | Tokens Added (lower is better) |
|---|---|---|---|---|
| Transform (Best) | 0.866 | 0.853 | 0.878 | 0.050 |
| Unstructured OSS | 0.426 | 0.498 | 0.715 | 0.119 |
OSS: unstructured.io/benchmarks (1,000+ pages). Transform Best: SCOREBench, 224 pages, Sept 2026.
:eight_pointed_black_star: Quick Start
There are several ways to use the unstructured library:
Run the library in a container or
Install the library
1. Install from PyPI
2. Install for local development
For installation with conda on Windows system, please refer to the documentation
Run the library in a container
The following instructions are intended to help you get up and running using Docker to interact with unstructured.
See here if you don't already have docker installed on your machine.
NOTE: we build multi-platform images to support both x86_64 and Apple silicon hardware. docker pull should download the corresponding image for your architecture, but you can specify with --platform (e.g. --platform linux/amd64) if needed.
We build Docker images for all pushes to main. We tag each image with the corresponding short commit hash (e.g. fbc7a69) and the application version (e.g. 0.5.5-dev1). We also tag the most recent image with latest. To leverage this, docker pull from our image repository.
bash
docker pull downloads.unstructured.io/unstructured-io/unstructured:latest
Once pulled, you can create a container from this image and shell to it.
bash
create the container
docker run -dt --name unstructured downloads.unstructured.io/unstructured-io/unstructured:latest
this will drop you into a bash shell where the Docker image is running
docker exec -it unstructured bash
You can also buiTop issues
#3718broken inference source code for 'hi_res', AttributeError: 'list' object has no attribute 'element_coords', the same code worked with previous versions of unstructured19 comments · 2024-10-13#2916bug/Execution speed is very slow in AWS LAMBDA environment15 comments · 2024-04-22#2631feat/Use local model for hi_res partition16 comments · 2024-03-11#3358bug/text-as-html-missing-content11 comments · 2024-07-08#3072Add manual coordinate constraints to `partition_pdf()`.9 comments · 2024-05-22#2976Lightweight installation unstructured[pdf] ?????9 comments · 2024-05-07#1038feat/ enable partition_json to partition any json9 comments · 2023-08-03#3865bug/<raise BadZipFile>8 comments · 2025-01-15
Discussions
#4490Rust-based document parser as a lightweight alternative in parsing pipelines (Founder's Lock)0 comments · 2026-09-18#3986Getting different output in text and text_as_html for table element1 comments · 2025-04-14#3972Using workflows for single files on user request0 comments · 2025-03-31#3970Add HTML Page Title to Element Metadata in partition_html()0 comments · 2025-03-28#3929[SSL: CERTIFICATE_VERIFY_FAILED]1 comments · 2025-02-19#3832'open_filename' from 'pdfminer.utils error0 comments · 2024-12-15#3805HTML Parser V2 Ontology?0 comments · 2024-12-03#3803Improper chunking for pdf1 comments · 2024-12-02
Latest release
0.27.37 0.27.37 2026-11-06What's Changed Fixes fix(lang): keep the languages a caller passes on short ASCII text by @chrikrah in https://github.com/Unstructured-IO/unstructured/pull/4601 fix(pdf): extract vector-heavy PDFs with the fast strategy instead of returning nothing by @CyMule in https://github.com/Unstructured-IO/unstructured/pull/4540 fix(cleaners): stop swapping en and em dashes in replace_unicode_quotes by @chrikrah in https://github.com/Unstructured-IO/unstructured/pull/4654 Other changes chore: start 0.27.37-dev after 0.27.36 release by @ryannikolaidis in https://github.com/Unstructured-IO/uns