Newsletter
Join the Community
Subscribe to our newsletter for the latest news and updates
MIT topic-to-article engine
STORM is an MIT topic-to-article engine: it searches the web, asks multi-perspective questions, then writes a long cited draft. Use it when the job is a Wikipedia-like first draft from a topic, not when you need a paper-screening table.
What makes it unique: the cover is STORM’s official demo article workspace — table of contents, generated draft, and references — from frontend/demo_light/assets/article_display.jpg in the Stanford repo.
pip install knowledge-storm (v1.1.0 adds LiteLLM).Best fit: writers and researchers who want a cited outline-to-draft, and can bring search/LLM keys to self-host.
Verified 2026-10-22 (Asia/Shanghai) from the stanford-oval/storm README and GitHub license.
Written path: README plus the hosted research preview and project site.
https://storm.genie.stanford.edu
https://github.com/stanford-oval/storm
Catalog row curated for AI Tools Nav. Cover is STORM’s official Streamlit demo article workspace (article_display.jpg): outline + generated article + references. License MIT; software $0; hosted preview $0.
2026-10-22 (Asia/Shanghai) editorial check.