[Bug]: Knowledge base retrieval function failure: Error invoking remote method 'knowledge-base:search': Error: 400 The parameter is invalid.

[ComradePenguin-1917]

Note

This issue was translated by Claude.

Issue Checklist

• I understand that issues are for feedback and problem solving, not for complaining in the comment section, and will provide as much information as possible to help solve the problem.
• My issue is not listed in the FAQ.
• I've looked at pinned issues and searched for existing Open Issues, Closed Issues, and Discussions, no similar issue or discussion was found.
• I've filled in short, clear headings so that developers can quickly identify a rough idea of what to expect when flipping through the list of issues. And not "a suggestion", "stuck", etc.
• I've confirmed that I am using the latest version of Cherry Studio.

Platform

Windows

Version

v1.7.13

Bug Description

I selected Qwen/Qwen3-Embedding-8B as the embedding model to build a knowledge base, without making any changes to dimensions or other settings. When I enabled this knowledge base in a conversation, an error occurred:

Steps To Reproduce

1. Create a knowledge base in the knowledge base interface and select Qwen/Qwen3-Embedding-8B as the embedding model
2. Enable this knowledge base in the conversation interface
3. Chat with the large model, and then an error occurs. Tested with GLM-4.7 and Deepseek-reasoner, both produced the same error

Expected Behavior

Correctly retrieve content from the knowledge base

Relevant Log Output

Additional Context

No response

---

Original Content

Issue Checklist

• I understand that issues are for feedback and problem solving, not for complaining in the comment section, and will provide as much information as possible to help solve the problem.
• My issue is not listed in the FAQ.
• I've looked at pinned issues and searched for existing Open Issues, Closed Issues, and Discussions, no similar issue or discussion was found.
• I've filled in short, clear headings so that developers can quickly identify a rough idea of what to expect when flipping through the list of issues. And not "a suggestion", "stuck", etc.
• I've confirmed that I am using the latest version of Cherry Studio.

Platform

Windows

Version

v1.7.13

Bug Description

选择使用Qwen/Qwen3-Embedding-8B作为嵌入模型构建一个知识库，没有进行任何维度或者其他设置的更改。在对话中启用该知识库，出现错误：

Steps To Reproduce

1. 在知识库界面创建知识库，选择Qwen/Qwen3-Embedding-8B作为嵌入模型
2. 在对话界面中启用该知识库
3. 和大模型进行对话，然后出错。试验了GLM-4.7和Deepseek-reasoner，均出现相同错误

Expected Behavior

正确检索知识库中的内容

Relevant Log Output

Additional Context

No response

[eeee0717]

Note

This comment was translated by Claude.

Which provider's embedding model?

---

Original Content

什么服务商的嵌入模型

[ComradePenguin-1917]

Note

This comment was translated by Claude.

Which provider's embedding model?

SiliconFlow's Qwen/Qwen3-Embedding-8B. Additionally, Zhipu's Embedding-3 also reports a different error:

---

Original Content

Which provider's embedding model?

硅基流动的Qwen/Qwen3-Embedding-8B，另外智谱的Embedding-3也会报另外的错：

[eeee0717]

Note

This comment was translated by Claude.

Looking at the error, it seems the input entity is too large. Could it be that the input is too long? If you create a new topic and ask a simple question, does it still report an error?

---

Original Content

看报错是输入实体太大了，不知道是不是input 太长了？创建一个新话题，简单的提问都会报错？

[ComradePenguin-1917]

Note

This comment was translated by Claude.

It might be because the length is too long; new topics can use the knowledge base normally.

---

Original Content

可能是因为长度太长，新话题能正常调用知识库