PandaWiki
A self-hosted AI knowledge base for documentation, with AI search, writing assistance, and integrations for website widgets and chatbots.

PandaWiki is a self-hosted knowledge base for teams publishing product guides, technical documentation, FAQs, or blogs. It gives readers AI-assisted ways to find answers within those documents.
- AI answers and search: Readers can ask questions about the knowledge base or use AI search, while authors can use AI assistance to draft content.
- Document editing and export: A rich text editor supports Markdown and HTML, with exports to Word, PDF, and Markdown.
- Content imports: Bring in material from web pages, website sitemaps, RSS feeds, and local files.
- Widgets and chatbots: Put the knowledge base on another website as a widget or connect it to chatbots in DingTalk, Feishu, and WeCom.
- Separate wiki sites: Manage collections of documents as distinct knowledge bases, each with its own wiki website.
PandaWiki is open source under AGPL-3.0 and runs on a Linux server with Docker. The recommended server has a two-core CPU, 4 GB of RAM, and 40 GB of disk space. Baizhi Cloud also offers a hosted SaaS edition.
AI features require a connected chat model. PandaWiki supports model APIs from Baizhi Cloud, DeepSeek, and OpenAI, while embedding and reranking models are built in. Self-hosting the wiki doesn't make cloud model calls local: requests to those providers leave your server.