The Qwen2.5 model deployed with vllm/sglang does not work with tools #84

Closed
opened 2026-02-16 10:18:01 -05:00 by yindo · 3 comments
Owner

Originally created by @Jimmy-L99 on GitHub (Mar 15, 2025).

Self Checks

  • This is only for bug report, if you would like to ask a question, please head to Discussions.
  • I have searched for existing issues Dify issues & Dify Official Plugins, including closed ones.
  • I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
  • [FOR CHINESE USERS] 请务必使用英文提交 Issue,否则会被关闭。谢谢!:)
  • Please do not modify this template :) and fill in all the required fields.

Dify version

0.15.2

Plugin version

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

I create an Agent in dify and use some tools such as get current time, get_weather and bar_chart, Qwen2.5-32B was deployed by vLLM or Sglang, but it cannot use the tools to make the right action.

By the way, in the vllm and sglang, I have already tried --tool-call-parser qwen or hermes.

Is there anything I'm missing?

✔️ Error log

No response

Originally created by @Jimmy-L99 on GitHub (Mar 15, 2025). ### Self Checks - [x] This is only for bug report, if you would like to ask a question, please head to [Discussions](https://github.com/langgenius/dify/discussions/categories/general). - [x] I have searched for existing issues [Dify issues](https://github.com/langgenius/dify/issues) & [Dify Official Plugins](https://github.com/langgenius/dify-official-plugins/issues), including closed ones. - [x] I confirm that I am using English to submit this report (我已阅读并同意 [Language Policy](https://github.com/langgenius/dify/issues/1542)). - [x] [FOR CHINESE USERS] 请务必使用英文提交 Issue,否则会被关闭。谢谢!:) - [x] Please do not modify this template :) and fill in all the required fields. ### Dify version 0.15.2 ### Plugin version * ### Cloud or Self Hosted Self Hosted (Docker) ### Steps to reproduce I create an Agent in dify and use some tools such as get current time, get_weather and bar_chart, Qwen2.5-32B was deployed by vLLM or Sglang, but it cannot use the tools to make the right action. By the way, in the vllm and sglang, I have already tried --tool-call-parser qwen or hermes. Is there anything I'm missing? ### ✔️ Error log _No response_
yindo added the bug label 2026-02-16 10:18:01 -05:00
yindo closed this issue 2026-02-16 10:18:01 -05:00
Author
Owner

@wertyac commented on GitHub (Mar 16, 2025):

I have the same issue. Can we use ReAct method? How to configure?

@wertyac commented on GitHub (Mar 16, 2025): I have the same issue. Can we use ReAct method? How to configure?
Author
Owner

@Jimmy-L99 commented on GitHub (Mar 17, 2025):

I have the same issue. Can we use ReAct method? How to configure?

You can enter the OpenAI-API-compatible---model setting and set the Function calling to Tool Call Instead of Function Call, then default tool can be use.
But the tool published with the workflow could only answer two rounds of conversations with not 100% call, and the third was unresponsive. So I prefer the chatflow.

@Jimmy-L99 commented on GitHub (Mar 17, 2025): > I have the same issue. Can we use ReAct method? How to configure? You can enter the `OpenAI-API-compatible`---`model setting` and set the `Function calling` to `Tool Call` Instead of `Function Call`, then default tool can be use. But the `tool` published with the `workflow` could only answer two rounds of conversations with not 100% call, and the third was unresponsive. So I prefer the `chatflow`.
Author
Owner

@dosubot[bot] commented on GitHub (Aug 16, 2025):

Hi, @Jimmy-L99. I'm Dosu, and I'm helping the dify-official-plugins team manage their backlog and am marking this issue as stale.

Issue Summary:

  • You reported that deploying the Qwen2.5-32B model with vLLM or Sglang in Dify 0.15.2 fails to properly use tools like get current time, get_weather, and bar_chart.
  • The issue persisted even when using --tool-call-parser options qwen and hermes.
  • Another user confirmed the problem and inquired about using the ReAct method.
  • You found that configuring the model setting to "Tool Call" instead of "Function Call" helps with default tool usage.
  • Using chatflow instead of workflow-published tools improves responsiveness and mitigates the issue.

Next Steps:

  • Please confirm if this issue is still relevant with the latest version of dify-official-plugins, and if so, you can keep the discussion open by commenting here.
  • Otherwise, I will automatically close this issue in 5 days.

Thank you for your understanding and contribution!

@dosubot[bot] commented on GitHub (Aug 16, 2025): Hi, @Jimmy-L99. I'm [Dosu](https://dosu.dev), and I'm helping the dify-official-plugins team manage their backlog and am marking this issue as stale. **Issue Summary:** - You reported that deploying the Qwen2.5-32B model with vLLM or Sglang in Dify 0.15.2 fails to properly use tools like get current time, get_weather, and bar_chart. - The issue persisted even when using --tool-call-parser options qwen and hermes. - Another user confirmed the problem and inquired about using the ReAct method. - You found that configuring the model setting to "Tool Call" instead of "Function Call" helps with default tool usage. - Using chatflow instead of workflow-published tools improves responsiveness and mitigates the issue. **Next Steps:** - Please confirm if this issue is still relevant with the latest version of dify-official-plugins, and if so, you can keep the discussion open by commenting here. - Otherwise, I will automatically close this issue in 5 days. Thank you for your understanding and contribution!
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langgenius/dify-official-plugins#84