PandaWiki
Open-source, self-hosted AI wiki from Chaitin — 9.9k GitHub stars
Overview
PandaWiki is an open-source, self-hosted knowledge-base system from Chinese cybersecurity firm Chaitin (长亭科技), combining a Markdown/HTML editor with RAG-backed AI Q&A, semantic search and embeddable chat widgets. It's Docker-deployable, AGPL-3.0 licensed, and has 9.9k+ GitHub stars.
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 These figures reflect RECATOOLS research as of 12 Jul 2026 and may be out of date or incomplete.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 PandaWiki
- Self-hosted via Docker (Linux)
- RAG-backed AI Q&A and semantic search
- Markdown/HTML rich-text editing with Word/PDF export
- Content ingestion from URLs, sitemaps, RSS, files
- Embeddable chat widget + DingTalk/Feishu/WeChat Work integration
- Bring-your-own LLM (DeepSeek-V4-Flash / BGE-M3 recommended)
ASEAN Perspective
PandaWiki 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).
For teams that want an AI-native docs and FAQ system without a SaaS subscription or sending everything to one model vendor, PandaWiki is a solid pick: you bring your own LLM (Chaitin recommends DeepSeek-V4-Flash for chat, BGE-M3 for embeddings), self-host via Docker, and keep the data. Nearly 10,000 GitHub stars and 986 forks since its mid-2025 open-sourcing suggest real developer adoption rather than a marketing repo, and the backing company, Chaitin, is a well-funded, established cybersecurity vendor rather than an unknown startup.
The trade-offs: documentation and the primary user community are almost entirely Chinese, so English-speaking teams will be piecing things together from translated blog posts and GitHub issues. And the AGPL-3.0 license means any org that modifies PandaWiki and offers it as a hosted service to others has to release those changes too — worth flagging to legal before deploying it that way.
What people say
Chaitin Tech (长亭科技), a Beijing cybersecurity company founded in 2014 by members of Tsinghua's Blue Lotus CTF team, is the company behind PandaWiki — the same team that placed third at Pwn2Own and drew Wall Street Journal coverage for breaking Windows, Mac and Linux targets in competition. Chaitin has raised well over 10 billion yuan across funding rounds, including a 2024 spin-off round described as the largest single raise by a Chinese cybersecurity company in years, plus a further 500 million yuan from the National AI Industry Investment Fund in January 2026 — so PandaWiki isn't a side project from a cash-strapped startup, it's a security vendor's bet on AI-native documentation tooling.
The project open-sourced in mid-2025 after what Chaitin describes as roughly a month of internal testing, and has grown to 9.9k GitHub stars, 986 forks, and a regular release cadence (v3.86.2 shipped June 29, 2026) — genuinely active for a Chinese open-source project, not an abandoned repo. It's built on a Go backend with a TypeScript frontend, licensed AGPL-3.0, and installs via Docker on Linux.
Functionally, PandaWiki covers the basics expected of a modern knowledge base — Markdown/HTML rich-text editing, export to Word/PDF/Markdown, content ingestion from URLs, sitemaps, RSS feeds and offline files — plus the AI layer that differentiates it: LLM-assisted drafting, RAG-backed Q&A and semantic search, and an embeddable widget plus chatbot integrations for DingTalk, Feishu and WeChat Work, the three enterprise chat platforms that matter in China. You supply your own model; Chaitin's docs recommend DeepSeek-V4-Flash for chat, BGE-M3 for embeddings and BGE-reranker-v2-m3 for reranking, though any OpenAI-compatible endpoint should work.
The main friction for an ASEAN or Western team is language: the documentation, community discussion (largely on CSDN, Zhihu and Chaitin's own forum) and default chat-platform integrations all assume a Chinese-language, Chinese-tooling environment. There's no G2 or Capterra listing — expected for a self-hosted open-source tool — so the GitHub activity itself is the best available adoption signal, and it's a genuinely strong one for a project barely a year old.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including PandaWiki's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with PandaWiki 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 PandaWiki directly →
Spotted something out of date? Suggest an update →
PandaWiki in the news
More in Productivity
