AIXtract

Enterprise-grade document extraction for LLM and NLP pipelines.

Installation

pip install aixtract
pip install "aixtract[all]"  # All format support

Quick Start

from aixtract import extract

result = extract("document.pdf")
print(result.content_markdown)

Supported Formats

Converter	Extensions	Dependencies
txt	.txt, .md, .rst, .log	none
csv	.csv, .tsv	none
json	.json	none
xml	.xml	none
archive	.zip	none
pdf	.pdf	pypdf, pdfplumber
docx	.docx, .doc	docx
xlsx	.xlsx, .xls	openpyxl │
pptx	.pptx, .ppt	pptx
html	.html, .htm	bs4
epub	.epub	ebooklib, bs4
image	.png, .jpg, .jpeg, .tiff, .bmp	PIL, pytesseract
audio	.mp3, .wav, .m4a, .flac, .ogg	whisper

Development Setup

If you are developing or contributing to aixtract, you can use the provided Makefile or pip directly.

Option 1: Using make (Recommended)

The Makefile is configured to automatically use the python and pip binaries inside .venv, so you don't even need to activate the environment to install.

Install for usage:

make install

Install for development (includes testing/linting tools):

make dev

Option 2: Using pip manually

If you prefer to run commands manually or don't have make installed:

1. Activate the virtual environment:

   source .venv/bin/activate

2. Install the project in editable mode:

   # For basic usage
   pip install -e .
   
   # For development (with dev dependencies)
   # NOTE: Quotes are required in zsh
   pip install -e ".[all,dev]"

Acknowledgments

AIXtract incorporates adapted components from CAMEL-AI, licensed under the Apache License 2.0. See NOTICE for details.

License

Apache License 2.0