AI · GitHub repo
STORM
AI knowledge curation system generating cited Wikipedia-style long-form reports from search.
STORM is an open-source research engine developed by Stanford University that automates the pre-writing and synthesis phases of long-form reports. It discovers topic perspectives, simulates expert-to-writer conversations grounded in search results, and writes structured, fully cited Wikipedia-style articles. Its collaborative variant, Co-STORM, allows human writers to guide the research discourse alongside AI agents.
What it's for
- Draft comprehensive Wikipedia-style overview articles with full reference citations
- Automate background research and topic exploration during the pre-writing stage
- Collaborate with AI agents using interactive mind maps to direct research topics
- Ground deep-dive research reports on private documents using custom vector databases
Features
- Perspective-guided question asking based on similar topic patterns
- Simulated multi-turn conversations between agent experts and topic writers
- Automatic hierarchical outline generation and multi-section article drafting
- Integrated reference tracking with inline academic-style citations
- Support for multiple search engines (Bing, You, DuckDuckGo, Brave, Serper, Tavily)
- Document grounding using VectorRM alongside live web retrieval
- LiteLLM compatibility for swapping model providers across different workflow stages
- Streamlit-based lightweight demo interface for local development
Pricing
Open source. Free and open-source (MIT license); requires user-supplied LLM and search API keys
Details
- Type
- GitHub repo
- Category
- AI
- Topic
- Research & Synthesis
- Open source
- Yes
- Platforms
- Code library, Command line, Web, Self-hosted
- For
- Researchers, writers, Wikipedia editors, and developers building automated research workflows.
- Alternative to
- Perplexity, Consensus
- Tags
- research, llm, wikipedia, citations, multi agent, stanford
Why it was saved
- @theaisurfer: Automates the labor-intensive knowledge curation and pre-writing stages of research, creating structured outlines and detailed referenced articles in minutes.
Compare
Related
AI-powered search engine providing conversational answers with cited sources.
Multi-agent system querying multiple LLMs that peer-review answers into one synthesis.
Python framework for building multi-modal AI agents with memory and tools
Conversational AI assistant for writing, research, reasoning, and coding
Browser that installs and runs open-source AI apps with one click
A comprehensive guide to prompt engineering techniques and LLM parameters by Google.
Updated