~/tools/growi
tool

GROWI

GROWI is an open source team wiki and knowledge base you can self-host — a Confluence and Notion alternative that stores docs as Markdown, with real-time editing, diagrams, full-text search, and group permissions.

What is GROWI?

GROWI is an open source wiki and knowledge base platform that stores your team’s documentation as Markdown pages in a hierarchical tree. Built by GROWI, Inc. and released under the MIT license, it adds real-time collaborative editing, draw.io diagrams, full-text search, and group-based permissions on top of plain Markdown.

What is GROWI best for?

Teams that want to keep manuals, meeting minutes, project notes, and reference material in one searchable place — and want to own the data on their own servers instead of a SaaS wiki. Its hierarchical page tree and Markdown-first editing suit engineering and operations teams comfortable with Markdown, especially anyone standardizing internal documentation across departments.

What can GROWI do?

  • Write pages in Markdown with a live editor, code highlighting, and formula support
  • Edit the same page with multiple people at once, with automatic saving and full revision history
  • Draw and embed diagrams inline with draw.io integration
  • Search across page contents and uploaded file attachments (full-text search via Elasticsearch)
  • Control access with user groups and page-level permissions
  • Authenticate through LDAP / Active Directory, SAML single sign-on, and OAuth (GitHub, Google)
  • Connect to Slack and Mattermost, plus IFTTT, and extend the app with npm plugins
  • Answer questions over your own docs with the AI Knowledge Assistant and Editor Assistant (RAG powered by OpenAI)
  • Store uploaded assets on the local filesystem, AWS S3, Google Cloud Storage, or MongoDB GridFS

Is GROWI free?

Yes — GROWI is fully open source under the MIT license and free to self-host; you only pay for the server it runs on. If you’d rather not run it yourself, the company offers GROWI.cloud, a paid managed hosting service billed at a fixed monthly rate per plan (with a limited free tier), so you’re paying for the hosting, not the software.

Where does GROWI fall short?

  • Full-text search needs a separate Elasticsearch service. MongoDB alone runs the app, but searching inside page contents and attachments requires a standalone Elasticsearch instance (7.x/8.x/9.x) with the kuromoji and ICU analysis plugins — a heavier dependency than most lightweight wikis need.
  • Documentation and community skew Japanese-first. GROWI originates in Japan; English docs and community support exist but can lag behind the Japanese side, so you may occasionally hit Japanese-only guidance.
  • The AI assistant depends on OpenAI. The Knowledge and Editor assistants send content to OpenAI via your own API key, so those features aren’t self-contained or offline — a consideration if data residency is the reason you self-hosted.

What does GROWI replace?

GROWI is a self-hosted alternative to Confluence and Notion. It covers the same team-wiki job — shared docs, search, permissions, and collaborative editing — but you run it on your own infrastructure and keep the data in-house, with no per-seat SaaS subscription.

FAQ

Is GROWI open source? Yes. GROWI is released under the MIT license, one of the most permissive open source licenses, so you can self-host, modify, and use it commercially without restriction.

Can I self-host GROWI for free? Yes. Self-hosting is free — the usual path is Docker Compose — and you only pay for your own server. GROWI.cloud is the optional paid managed offering.

Is GROWI a good Confluence alternative? For teams that want Markdown-first documentation on their own servers, yes. It matches Confluence’s core wiki, search, and permission features without per-user pricing. If you prefer a fully hosted tool with polished onboarding, Confluence Cloud or Notion may feel more turnkey.

What do I need to run GROWI? A server with Docker, Node.js, and MongoDB. Add a Redis cache and an Elasticsearch instance if you want full-text search; both are optional but search depends on Elasticsearch.