Parallel LLM nodes in the workflow cannot stream output results. #13307

Closed
opened 2026-02-21 19:11:34 -05:00 by yindo · 4 comments
Owner

Originally created by @qazxswaaa on GitHub (Apr 24, 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 search for existing issues, 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

1.0

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

在工作流中通过多个if-else节点进行判断,使得后续可以多个LLM节点同时运行,但是在运行之后结果只会显示一个结束节点的输出内容,在使用API进行流式调用时,LLM节点的流式思考过程也没有了

Image

✔️ Expected Behavior

No response

Actual Behavior

No response

Originally created by @qazxswaaa on GitHub (Apr 24, 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 [search for existing issues](https://github.com/langgenius/dify/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 1.0 ### Cloud or Self Hosted Self Hosted (Docker) ### Steps to reproduce 在工作流中通过多个if-else节点进行判断,使得后续可以多个LLM节点同时运行,但是在运行之后结果只会显示一个结束节点的输出内容,在使用API进行流式调用时,LLM节点的流式思考过程也没有了 ![Image](https://github.com/user-attachments/assets/7b183432-b832-4d2b-96cb-4c599b5d7fc2) ### ✔️ Expected Behavior _No response_ ### ❌ Actual Behavior _No response_
yindo added the 🐞 bug🌊 feat:workflow labels 2026-02-21 19:11:34 -05:00
yindo closed this issue 2026-02-21 19:11:34 -05:00
Author
Owner

@dosubot[bot] commented on GitHub (Apr 24, 2025):

Hello @qazxswaaa! I'm Dosu, a bot here to help you solve bugs, answer questions, and assist you in becoming a contributor. I'm ready to help while you wait for a human maintainer. How can I assist you today?

I found similar threads discussing issues with streaming output in parallel LLM nodes within Dify workflows:

  1. Streaming Output Unavailability in LLM Node - This discussion is open and does not have a solution yet [1].
  2. Non-Streaming Output from LLM Model in Dify Usage - This discussion is also open and unsolved [2].

Regarding your issue, Dify's current limitations include the inability to stream output from variable aggregator nodes, which is expected behavior. Additionally, complex workflows involving conditional branches may switch to blocking mode at the last LLM node. This can be resolved by applying the patch from PR #12355, which addresses streaming output issues in Dify versions 0.14.1 and 0.14.2 [3][4].

For parallel execution of LLM nodes, there is no specific information available, but it might be related to the limitations mentioned above [3].

To continue talking to Dosu, mention @dosu.


Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other

@dosubot[bot] commented on GitHub (Apr 24, 2025): <!-- Greeting --> Hello @qazxswaaa! I'm [Dosu](https://dosu.dev), a bot here to help you solve bugs, answer questions, and assist you in becoming a contributor. I'm ready to help while you wait for a human maintainer. How can I assist you today? <!-- Answer --> I found similar threads discussing issues with streaming output in parallel LLM nodes within Dify workflows: 1. [Streaming Output Unavailability in LLM Node](https://github.com/langgenius/dify/discussions/16521) - This discussion is open and does not have a solution yet <sup>[[1]](https://github.com/langgenius/dify/discussions/16521)</sup>. 2. [Non-Streaming Output from LLM Model in Dify Usage](https://github.com/langgenius/dify/discussions/12393) - This discussion is also open and unsolved <sup>[[2]](https://github.com/langgenius/dify/discussions/12393)</sup>. Regarding your issue, Dify's current limitations include the inability to stream output from variable aggregator nodes, which is expected behavior. Additionally, complex workflows involving conditional branches may switch to blocking mode at the last LLM node. This can be resolved by applying the patch from PR #12355, which addresses streaming output issues in Dify versions 0.14.1 and 0.14.2 <sup>[[3]](https://github.com/langgenius/dify/issues/16647)</sup><sup>[[4]](https://github.com/langgenius/dify/issues/11985)</sup>. For parallel execution of LLM nodes, there is no specific information available, but it might be related to the limitations mentioned above <sup>[[3]](https://github.com/langgenius/dify/issues/16647)</sup>. <!-- Dosu Comment Footer --> *To continue talking to [Dosu](https://dosu.dev), mention @dosu.* --- <sup>Help Dosu learn! Give it feedback: [Great Response](https://app.dosu.dev/response-feedback/02b9f5cb-059a-4a45-9aea-99c3a334902a?feedback_type=great_response) | [Irrelevant Answer](https://app.dosu.dev/response-feedback/02b9f5cb-059a-4a45-9aea-99c3a334902a?feedback_type=irrelevant_answer) | [Incorrect Sources](https://app.dosu.dev/response-feedback/02b9f5cb-059a-4a45-9aea-99c3a334902a?feedback_type=incorrect_sources) | [Too Verbose](https://app.dosu.dev/response-feedback/02b9f5cb-059a-4a45-9aea-99c3a334902a?feedback_type=too_verbose) | [Hallucination](https://app.dosu.dev/response-feedback/02b9f5cb-059a-4a45-9aea-99c3a334902a?feedback_type=hallucination) | [Bug Report](https://app.dosu.dev/response-feedback/02b9f5cb-059a-4a45-9aea-99c3a334902a?feedback_type=bug_report) | [Other](https://app.dosu.dev/response-feedback/02b9f5cb-059a-4a45-9aea-99c3a334902a?feedback_type=other)</sup>
Author
Owner

@crazywoola commented on GitHub (Apr 24, 2025):

[FOR CHINESE USERS] 请务必使用英文提交 Issue,否则会被关闭。谢谢!:)

@crazywoola commented on GitHub (Apr 24, 2025): [FOR CHINESE USERS] 请务必使用英文提交 Issue,否则会被关闭。谢谢!:)
Author
Owner

@crazywoola commented on GitHub (Apr 24, 2025):

Please upgrade to latest version, this has been fixed already.

@crazywoola commented on GitHub (Apr 24, 2025): Please upgrade to latest version, this has been fixed already.
Author
Owner

@qazxswaaa commented on GitHub (Apr 25, 2025):

Please upgrade to latest version, this has been fixed already.

Version 1.3 has been updated, but the LLM nodes running in parallel still have no streaming output. No text_chunk response when the interface is invoked.

Image

@qazxswaaa commented on GitHub (Apr 25, 2025): > Please upgrade to latest version, this has been fixed already. Version 1.3 has been updated, but the LLM nodes running in parallel still have no streaming output. No text_chunk response when the interface is invoked. ![Image](https://github.com/user-attachments/assets/3fa7629c-c48d-4b84-b43f-c3aa686fc887)
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langgenius/dify#13307