STORM: Search web data based on Topic to generate papers with citations, long paper reports

General Introduction

STORM is a knowledge integration and article generation system developed by the Oval team at Stanford University. It focuses on generating exhaustive Wikipedia-like articles (systematic papers) from scratch. The system utilizes large-scale language models for topic research, prepares summaries and simulates expert conversations from actual Internet sources to support writers in asking questions to produce lengthy articles containing citations.

All you have to do is type in a topic and it will search for information from hundreds of websites to write a comprehensive research article for you. From format/topic/content/citation/absolute expert level. The point is: it's completely free and open to the public! This is simply a strong challenger to Google Research, so give it a try!

STORM New homepage, Co-STORM multi-player discussion mode is also available

New version of the detail page

old version

Function List

Researching topics and integrating knowledge
Creating topic summaries
Simulated conversations based on Internet sources
Automated writing to generate Wikipedia-like articles
Follow-up on topic understanding and questions through expert dialog
Code available and explored at GitHub
Live demo available

Using Help

Check out the GitHub repository for installation and configuration steps
Running code for prewriting phase of topic research
Conducting the writing phase of article writing and refinement

STORM runs online

1. Select Topic
2. Outline generation
3. Searching the Internet for cited material
4. Generate long articles with citations step by step based on an outline and a dialog strategy that mimics Wikipedia's professional writers and editors.