STORM (Stanford)
Stanford's open-source tool that researches and drafts a cited report.
Overview
STORM is Stanford OVAL's open-source research pipeline: it simulates multi-perspective expert interviews to gather sources, builds an outline, then drafts a long-form, Wikipedia-style article with citations. Free to self-host from GitHub; a hosted demo exists but is not a commercial product.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
What you can produce with STORM (Stanford)
- Multi-perspective simulated expert interviews for research
- Automatic outline generation
- Full-length article drafting with inline citations
- Co-STORM human-in-the-loop research mode
- Self-hostable via GitHub (Python)
- Free public research demo
ASEAN Perspective
STORM (Stanford) in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
STORM is still the reference implementation for grounded, structure-first long-form generation -- its trick of simulating a Wikipedia-writer persona interviewing multiple expert personas before drafting anything produces noticeably better-organized, more broadly-sourced output than a single-pass RAG prompt. The GitHub repo has crossed 29,000 stars and stays active, with a Co-STORM mode (from the EMNLP 2024 paper) that lets a human steer the research conversation instead of only reading the final article.
It's a research artifact, not a SaaS product: you need to bring your own LLM and search API keys, run Python, and accept that the hosted genie.stanford.edu demo is explicitly a prototype, not a service with uptime guarantees. Stanford's own team documents source bias and occasional fact-misattribution as open problems. Treat output as a sourced first draft to edit, not publishable copy -- and it's the strongest free option in that lane.
What people say
STORM doesn't have consumer reviews in the G2/Capterra sense -- it's a research codebase, not a product with a pricing page -- so its reputation lives in GitHub activity, academic citation, and write-ups from people who've run it against their own document sets.
The underlying paper, 'Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models' by Yijia Shao, Yucheng Jiang, Theodore Kanell, Peter Xu, Omar Khattab and Monica Lam, was published at NAACL 2024 and got independent coverage from the Wikipedia Signpost itself -- notable because it's Wikipedia's own community newsletter reviewing a tool built to mimic Wikipedia's writing process. The GitHub repository (stanford-oval/storm) has passed 29,000 stars and continues to see active pull requests and issues into mid-2026, including ongoing work on retrieval modules and a documented request for a 'knowledge curation & memory failure modes' guide -- a sign the maintainers are still engaging with real usage problems rather than letting the repo go stale two years after launch.
Coverage from AI-tooling writers consistently frames STORM as one of the more credible open-source approaches to long-form, cited generation, specifically because the multi-perspective interview step forces the system to seek out sources it wouldn't find with a single search query. The Co-STORM follow-up, published at EMNLP 2024, adds a mode where a human can sit inside the research loop and redirect it mid-conversation instead of only reviewing the final draft -- reviewers who've tried both describe Co-STORM as noticeably more useful for real research work than the original hands-off version.
The tradeoffs are the ones Stanford's own team has written up openly: the system can lean on whichever sources rank highly in search results (source bias), and it can occasionally misattribute a claim to the wrong source during synthesis. There's no commercial API or support contract -- self-hosting means managing your own LLM and search provider costs, supplying your own retrieval backend, and keeping the codebase updated yourself as the repo evolves.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including STORM (Stanford)'s official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with STORM (Stanford) unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to STORM (Stanford) directly →
Spotted something out of date? Suggest an update →
STORM (Stanford) in the news
More in Research & Data
