Glossary
What is knowledge source?
One item in a knowledge base, such as a crawled website, an uploaded file or a connected Notion workspace, and often the unit a plan limits.
A knowledge source is one item in a knowledge base: a crawled website, a sitemap import, an uploaded PDF, a connected Notion workspace or Google Drive folder. It is the unit you add, refresh and delete, and often the unit a plan limits. Intercom and Zendesk call them knowledge sources; Tidio says data sources.
It matters because vendors count sources differently, and the count decides what a free or starter plan can hold. One vendor may count every crawled page as a source, another the whole crawl as one, a third counts characters or megabytes instead. A limit of "5 sources" can therefore mean five pages or five entire websites. Read the counting rule before comparing plans, and check what happens at the cap: refused imports, a prompt to upgrade, or silent truncation.
Where definitions split further: snippets and question-and-answer pairs sit inside the knowledge base but are often not counted as sources, because they are typed in the dashboard rather than imported. Published help center articles are usually mirrored in automatically and not counted either. The text they contain still counts toward any size cap, so a workspace can be under its source limit and over its text limit at the same time, and the second limit is the one that bites on a large site.
Related terms
- Knowledge baseThe private collection of pages, files and snippets an AI support agent searches before every reply, not the same thing as a public help center.
- CrawlAn import that starts at one address and follows a site's links to collect pages for the knowledge base, usually with a page and depth limit.
- Retrieval-augmented generation (RAG)Searching your content for passages relevant to a question and handing them to the model as it writes the answer.
See it in a real chat
PepoChat runs on this site. Ask the agent in the corner anything about the product.
