Creativity · Agent Protocol

Stanford STORM: Research Agent for Long-Form Articles

STORM (Synthesis of Topic Outlines through Retrieval and Multi-perspective question asking), published by Stanford OVAL in 2024, is an open-source research agent that drafts Wikipedia-style long-form articles. It simulates expert personas interviewing each other to surface angles a single chain-of-thought would miss, then organizes an outline and drafts cited sections. Co-STORM extends it to collaborative, human-in-the-loop mode.

Protocol facts

Sponsor
Stanford OVAL Lab
Status
stable
Spec
https://github.com/stanford-oval/storm
Interop with
OpenAI, Anthropic, You.com, Bing Search

Frequently asked questions

What's unique about the multi-perspective approach?

STORM simulates multiple expert personas asking each other questions on the topic. Each persona pulls different sources, which together surface a more complete set of angles than a single planner would.

What is Co-STORM?

A 2024 extension where a human joins the simulated panel as another 'expert', can redirect the conversation, and curates the emerging outline — bringing human judgment into the research loop.

Is STORM tied to one search engine?

No — connectors exist for You.com, Bing, Serper, DuckDuckGo, and others. The retrieval abstraction makes it easy to plug in enterprise corpora.

Sources

  1. STORM paper (NAACL 2024) — accessed 2026-04-20
  2. STORM GitHub — accessed 2026-04-20

Related

  • GPT Researcher: Autonomous Research Agent — GPT Researcher is an open-source autonomous research agent that drafts a plan, issues web queries across many sources, deduplicates, and writes a cited research report — all without a human in the loop.
  • Perplexity Deep Research — Perplexity Deep Research is an autonomous multi-step research agent that browses the web for several minutes, synthesizes dozens of sources, and writes a cited long-form report for a single prompt.
  • Sakana AI Scientist: Fully Automated Research Pipeline — AI Scientist by Sakana AI is an end-to-end agent pipeline that proposes ML research ideas, writes experiment code, runs experiments, analyzes results, and drafts a LaTeX paper — the first demonstration of fully autonomous ML research.
  • All Agent Protocols — Browse the complete agent protocols library
  • VSET — Programmes offered — B.Tech programs (CSE, AI&ML, AI&DS, Electronics, IIoT)