Extractium™ gathers what your organization already publishes into one compendium: a searchable collection of your content, written as static files. Point it at your website, knowledge base portal, GitHub repositories, YouTube channel, library repository, or a folder of files. It gathers the content, prepares it for both keyword and meaning-based search, and writes it out in several formats. AI assistants can read the result, small language models can search it inside a web app or a script, and none of it depends on any one AI provider or on a server you have to run.
Unlike a vector database, Extractium™ needs no server, no database, and no API to run. Every output is a static file that you can host anywhere, including GitHub Pages, and the same build feeds all of them at once. Sources and outputs are plug-ins, so you can add your own if the built-in ones do not cover your needs.
| Output | Files | Best for |
|---|---|---|
| Search index | compendium.json.gz, compendium-full.json.gz |
Fast keyword and meaning-based search with nothing to run: a search box on your website, a script, an AI assistant on your computer, or a hosted search endpoint. |
| llms.txt files | llms.txt, llms/ |
AI assistants and platforms that can read web pages but cannot call tools. llms.txt lists your sources, and each source has a short index file of its pages. Also a readable list of everything that was indexed. |
| SQLite database | compendium.sqlite |
SQL queries and reports, or loading the content into a hosted database. |
| Markdown folder | okf/ |
Reading and editing the content as ordinary files, sharing it with other tools that use the Open Knowledge Format, or feeding it into another Extractium™ build. |
Extractium™ grew out of the indexing engine in Field Station AI™, which remains an example front-end implementation for how to access the JSON output in JavaScript.
-
On Windows, download
extractium-<version>-windows-portable.zipfrom the releases page, unblock it in its Properties, unzip it into the folder where your compendium should live, and double-clickrun.bat. Nothing else is downloaded, and you need no admin rights and no Python. -
On macOS or Linux, or on Windows without the zip, download the installer, install.sh or install.bat, and run it. It installs Extractium™ for your account, with an
extractiumcommand and an entry in the Start menu or its equivalent, and asks IT for nothing:bash install.sh # macOS and Linux install.bat # Windows
-
Run
extractiumin the folder where your compendium should live, or open Extractium from the menu. It asks whether to set up in the browser or in the terminal. Either way it asks for the name of your compendium, a short name for its files, and the website to crawl; the browser also asks where the compendium should live. The terminal path then builds a first index limited to 25 pages; the browser path writes the settings file from the page, and the nextextractiumbuilds. -
Open
dist/llms.txt, then the file it links to underdist/llms/, to see which pages were indexed. When the list looks right, runextractiumagain to build the whole site. To change what is crawled, editconfig.yaml, or runextractium uito change it from a page in your browser. Seeexamples/config.efdc.yamlfor a complete example that uses every source type. -
To use a Python development environment instead, clone the repository, run
pip install -e ".[dev,code,youtube,whisper,pdf,keywords]", thenpython -m extractium.cli initto writeconfig.yamlandpython -m extractium.cli build --config config.yamlto build.
The zip carries the three models a build and a search use. Every other install downloads the embedding model, about 130 MB, on the first build, and later builds reuse it. How to install has the details, including what the installer does when your computer stops it.
- An overview and user guide is available in the EFDC Knowledge Base at: https://michmed.org/efdc-kb
- Detailed documentation, for users and developers, is available under docs/.
- FieldStationAI™: https://github.com/DepressionCenter/FieldStationAI
- Mobile Technologies Core, the group that develops and maintains Extractium™ and Field Station AI™.
- EFDC Knowledge Base, the documentation site referenced above.
The Mobile Technologies Core provides investigators across the University of Michigan the support and guidance needed to utilize mobile technologies and digital mental health measures in their studies. Experienced faculty and staff offer hands-on consultative services to researchers throughout the University – regardless of specialty or research focus.
Learn more at: https://depressioncenter.org/mobiletech.
To get in touch, contact the individual developers in the check-in history.
If you need assistance identifying a contact person, email the EFDC's Mobile Technologies Core at: efdc-mobiletech@umich.edu.
- FieldStationAI™: A research platform for mobile and digital mental health studies. Its crawling and indexing engine was extracted into this project. https://github.com/DepressionCenter/FieldStationAI
- BAAI/bge-small-en-v1.5: The sentence-embedding model used by every build and every client. MIT license. https://huggingface.co/BAAI/bge-small-en-v1.5
- llms.txt: The convention the
llms.txtoutput follows. https://llmstxt.org/ - Open Knowledge Format: The Markdown-with-front-matter format the
okfoutput writes and theokfsource reads. https://github.com/GoogleCloudPlatform/open-knowledge-format - Python libraries used: requests, curl_cffi, Beautiful Soup 4, Sentence Transformers, NumPy, Python-Markdown, PyYAML, and optionally Tree-sitter with its language grammars, Universal Ctags, youtube-transcript-api, pypdf, pytest, and uv.
- JavaScript libraries used by the examples: @huggingface/transformers and the ONNX Runtime it brings.
Every dependency license was checked for compatibility with the GNU General Public License v3.0 or later. The list, with each license, is in docs/compliance.md.
Copyright © 2026 The Regents of the University of Michigan
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program. If not, see https://www.gnu.org/licenses/gpl-3.0-standalone.html.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. You should have received a copy of the license included in the section entitled "GNU Free Documentation License". If not, see https://www.gnu.org/licenses/fdl-1.3-standalone.html
If you find this repository, code or paper useful for your research, please cite it.
Mongefranco, Gabriel (2026). Extractium™. University of Michigan. Software. https://github.com/DepressionCenter/extractium
DOI: 10.5281/zenodo.22754640
Copyright © 2026 The Regents of the University of Michigan