---
id: "b918"
language: "en"
library: "tools"
official_url: "https://github.com/Unstructured-IO/unstructured"
profile_updated_on: "2026-10-21"
---

# Unstructured

Open-source Python components partition PDFs, images and office documents into structured elements for LLM retrieval and knowledge pipelines.

- Company / organization: Unstructured Technologies, Inc.
- Company country / region: United States · Company headquarters
- Category: Model development
- Official website: [https://github.com/Unstructured-IO/unstructured](https://github.com/Unstructured-IO/unstructured)
- Profile updated: 2026-10-21


## Overview

This entry covers Unstructured’s open-source preprocessing library. File-specific parsers, layout and OCR components convert documents into structured elements for LLM retrieval and knowledge pipelines. It requires a development environment and format-dependent dependencies. The library differs from hosted Transform; hosted model results should not be attributed to it. Default lightweight telemetry can be disabled as documented.

## Company background

- Company website: https://unstructured.io/
- Business focus: Provides document transformation and preprocessing for AI applications, maintaining both open-source libraries and commercial hosted products.

## Official sources and references

- [Official website](https://github.com/Unstructured-IO/unstructured)
- [Documentation](https://github.com/Unstructured-IO/unstructured/blob/main/README.md)

## Reading & subscriptions

- [Interactive profile](../../../products/unstructured--b918.html)
- [Read](../../../read/en/products/unstructured--b918.html)
- [JSON](../../../read/en/products/unstructured--b918.json)

Descriptions summarize public information; consult official websites for features, pricing and regional availability.