> For the complete documentation index, see [llms.txt](https://docs.cherry-ai.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cherry-ai.com/cherry-studio/preview/knowledge-base.md).

# 知识库

知识库就像给 AI 配一本**专属参考书**：你把自己的文档、笔记、网址塞进去，之后聊天时让 AI 翻这本书来回答你的问题。

<figure><img src="/files/P7q1NPTTd9fvuCMJJvMP" alt=""><figcaption><p>知识库入口：左上 <code>+ 添加</code> 创建第一个知识库</p></figcaption></figure>

## 用知识库能干什么？

举几个真实场景：

* **公司知识助手**：把产品手册、API 文档、内部规范全塞进去，员工问问题时 AI 自动答
* **个人资料管家**：把你历年的工作笔记、读书摘录、邮件存档放进去，问 AI"我去年在哪个 PPT 里提过那个分析框架"
* **学习陪练**：把课件、论文塞进去，让 AI 帮你按章节出题、解答疑惑
* **合同/法规速查**：把法条、合同模板放进去，问 AI 具体条款的应用

## 为什么用知识库，不直接把文件丢给 AI？

直接丢文件的限制：

* 每次提问都要重新上传，麻烦
* 单次对话有长度限制，长文档塞不下
* 跨对话不能复用

**知识库解决了上面所有问题**：上传一次，之后任何对话都可调用，且能从大量资料里"精准抓取相关段落"喂给 AI。

## 怎么用？

* 第一次用：看 [完整知识库教程](/knowledge-base/knowledge-base.md)
* 想加图片 / 扫描 PDF：先看 [文档预处理](/knowledge-base/document-preprocessing.md)，让 AI 能"读懂"图片里的文字
* 想了解嵌入模型怎么选：看 [嵌入模型参考](/knowledge-base/emb-models-info.md)
* 想了解数据存哪：看 [知识库数据](/knowledge-base/data.md)

## 与其他能力的组合

* **知识库 + 助手**：给某个助手"挂载"知识库，它就专精这个领域
* **知识库 +** [**Cherry Agent**](/cherry-studio/preview/agent.md)：让 Agent 在任务过程中自己查知识库
* **知识库 +** [**频道**](/advanced-basic/agent-channels.md)：把"会查公司文档"的 Agent 派到飞书群里值班

> 推荐先阅读 [概念入门](/advanced-basic/concepts-101.md) 了解知识库与 Agent / MCP 等功能的关系。

***

### 💡 获取帮助与提交反馈

如果您在配置或使用过程中遇到任何疑问、Bug 或有功能改进建议，请参考 [反馈与建议](/question-contact/suggestions.md) 中提供的官方渠道。


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.cherry-ai.com/cherry-studio/preview/knowledge-base.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.