langgenius/dify
1
0
Fork 0
mirror of https://github.com/langgenius/dify.git synced 2026-07-21 12:05:30 -04:00
Code Issues 775 Packages Projects Releases 10 Wiki Activity

Streaming output issue #20350

New Issue
Closed
opened 2026-02-21 20:06:58 -05:00 by yindo · 3 comments
yindo commented 2026-02-21 20:06:58 -05:00
Owner
Copy Link
Copy Source

Originally created by @15171452351 on GitHub (Nov 16, 2025).

Self Checks

  • I have read the Contributing Guide and Language Policy.
  • 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, otherwise it will be closed.
  • 【中文用户 & Non English User】请使用英语提交，否则会被关闭 ：）
  • Please do not modify this template :) and fill in all the required fields.

Dify version

1.8.0 and 1.10.0

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

All other conditions being equal, I found that when calling the API of a regular chat assistant-type application, the output is returned in a typing-like streaming manner.

Image But when calling the API of an agent-type application, only the last agent_message returns the full content, rather than truly streaming output. Is this because the current agent application doesn't support it yet, or is it a bug? Image

✔️ Expected Behavior

The agent application should have truly streaming output responses.

❌ Actual Behavior

But unfortunately, the agent application interface returns all the content inside a single data field.

Originally created by @15171452351 on GitHub (Nov 16, 2025). ### Self Checks - [x] I have read the [Contributing Guide](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md) and [Language Policy](https://github.com/langgenius/dify/issues/1542). - [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, otherwise it will be closed. - [x] 【中文用户 & Non English User】请使用英语提交，否则会被关闭 ：） - [x] Please do not modify this template :) and fill in all the required fields. ### Dify version 1.8.0 and 1.10.0 ### Cloud or Self Hosted Self Hosted (Docker) ### Steps to reproduce All other conditions being equal, I found that when calling the API of a regular chat assistant-type application, the output is returned in a typing-like streaming manner. <img width="1143" height="586" alt="Image" src="https://github.com/user-attachments/assets/0a3c4bbb-c184-4c14-8c35-5e115fd7e93e" /> But when calling the API of an agent-type application, only the last agent_message returns the full content, rather than truly streaming output. Is this because the current agent application doesn't support it yet, or is it a bug? <img width="1146" height="612" alt="Image" src="https://github.com/user-attachments/assets/c84932f2-5162-4937-89d1-e56c55f9a12d" /> ### ✔️ Expected Behavior The agent application should have truly streaming output responses. ### ❌ Actual Behavior But unfortunately, the agent application interface returns all the content inside a single data field.
yindo added the 🐞 bug label 2026-02-21 20:06:58 -05:00
yindo closed this issue 2026-02-21 20:06:58 -05:00
yindo commented 2026-02-21 20:06:58 -05:00
Author
Owner
Copy Link
Copy Source

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

Hi @15171452351! I'm Dosu and I’m helping the dify team.

The lack of streaming output in agent-type applications is a known limitation in Dify, not just a bug. Agent nodes have historically not supported streaming output like LLM nodes do. Streaming support for agents is being improved, but it depends on your model and plugin declaring the right features (like stream-tool-call). Some regressions in agent streaming were present in v1.8.x and were fixed in v1.9.1 and later, but even in v1.10.0, streaming for agents is not universally available—it depends on your workflow, model, and plugin configuration. For example, the Ollama plugin does not stream in agent mode unless you manually add the required feature keys to its YAML config. Reverse proxy buffering (like nginx) and HTTP/2 can also block streaming even if everything else is set up correctly. There is an open issue tracking this, and official support is improving, but for now, workarounds include using LLM nodes, modifying plugin YAML, or adjusting workflow settings. For more details and troubleshooting steps, see this discussion and this issue.

To reply, just mention @dosu.

How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other  Chat with Dosu Join Discord Share on X

@dosubot[bot] commented on GitHub (Nov 16, 2025): <!-- Greeting --> Hi @15171452351! I'm [Dosu](https://go.dosu.dev/dosubot) and I’m helping the dify team. <!-- Answer --> The lack of streaming output in agent-type applications is a known limitation in Dify, not just a bug. Agent nodes have historically not supported streaming output like LLM nodes do. Streaming support for agents is being improved, but it depends on your model and plugin declaring the right features (like `stream-tool-call`). Some regressions in agent streaming were present in v1.8.x and were fixed in v1.9.1 and later, but even in v1.10.0, streaming for agents is not universally available—it depends on your workflow, model, and plugin configuration. For example, the Ollama plugin does not stream in agent mode unless you manually add the required feature keys to its YAML config. Reverse proxy buffering (like nginx) and HTTP/2 can also block streaming even if everything else is set up correctly. There is an open issue tracking this, and official support is improving, but for now, workarounds include using LLM nodes, modifying plugin YAML, or adjusting workflow settings. For more details and troubleshooting steps, see [this discussion](https://github.com/langgenius/dify/issues/20952) and [this issue](https://github.com/langgenius/dify/issues/27284). <!-- Dosu Comment Footer --> *To reply, just mention [@dosu](https://go.dosu.dev/dosubot).* --- <sup>How did I do? [Good](https://app.dosu.dev/response-feedback/6bb70540-54b6-455e-b1b7-5d982ec74632?feedback_type=great_response) | [Irrelevant](https://app.dosu.dev/response-feedback/6bb70540-54b6-455e-b1b7-5d982ec74632?feedback_type=irrelevant_answer) | [Incorrect](https://app.dosu.dev/response-feedback/6bb70540-54b6-455e-b1b7-5d982ec74632?feedback_type=incorrect_sources) | [Verbose](https://app.dosu.dev/response-feedback/6bb70540-54b6-455e-b1b7-5d982ec74632?feedback_type=too_verbose) | [Hallucination](https://app.dosu.dev/response-feedback/6bb70540-54b6-455e-b1b7-5d982ec74632?feedback_type=hallucination) | [Report 🐛](https://app.dosu.dev/response-feedback/6bb70540-54b6-455e-b1b7-5d982ec74632?feedback_type=bug_report) | [Other](https://app.dosu.dev/response-feedback/6bb70540-54b6-455e-b1b7-5d982ec74632?feedback_type=other)</sup>&nbsp;&nbsp;[![Chat with Dosu](https://dosu.dev/dosu-chat-badge.svg)](https://app.dosu.dev/6088a4f4-073a-43f3-b3d4-4f06fb0c05a0/ask?utm_source=github)&nbsp;[![Join Discord](https://img.shields.io/badge/join-5865F2?logo=discord&logoColor=white&label=)](https://go.dosu.dev/discord-bot)&nbsp;[![Share on X](https://img.shields.io/badge/X-share-black)](https://twitter.com/intent/tweet?text=%40dosu_ai%20helped%20me%20solve%20this%20issue!&url=https%3A//github.com/langgenius/dify/issues/28258)
yindo commented 2026-02-21 20:06:58 -05:00
Author
Owner
Copy Link
Copy Source

@Cursx commented on GitHub (Nov 19, 2025):

@15171452351 If your model is deployed through the Olama or Tongyi plugins, it should now be able to stream output in the agent module. You need to update the ollama plugin to version 0.0.8 (I submitted a related PR to implement streaming output and streaming tool calls).

Image
@Cursx commented on GitHub (Nov 19, 2025): @15171452351 If your model is deployed through the Olama or Tongyi plugins, it should now be able to stream output in the agent module. You need to update the ollama plugin to version 0.0.8 (I submitted a related PR to implement streaming output and streaming tool calls). <img width="1728" height="792" alt="Image" src="https://github.com/user-attachments/assets/b311c0dc-b480-4974-ab86-091daa65a1ea" />
yindo commented 2026-02-21 20:06:58 -05:00
Author
Owner
Copy Link
Copy Source

@Cursx commented on GitHub (Nov 19, 2025):

@15171452351 If your model is deployed through the Olama or Tongyi plugins, it should now be able to stream output in the agent module. You need to update the ollama plugin to version 0.0.8 (I submitted a related PR to implement streaming output and streaming tool calls).

Image

Whether it's the internal display of DIFY's agent module or API calls.

@Cursx commented on GitHub (Nov 19, 2025): > [@15171452351](https://github.com/15171452351) If your model is deployed through the Olama or Tongyi plugins, it should now be able to stream output in the agent module. You need to update the ollama plugin to version 0.0.8 (I submitted a related PR to implement streaming output and streaming tool calls). > > <img alt="Image" width="1728" height="792" src="https://private-user-images.githubusercontent.com/33718736/516575359-b311c0dc-b480-4974-ab86-091daa65a1ea.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.VQbvrRUXlAQJKR5Z2vpiuosYu805PlAGgDhkaoIgXbg"> Whether it's the internal display of DIFY's agent module or API calls.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Knowledge Pipeline
Proposal
WIP
cloud
dependencies
feat:webapp
good first issue
outdated
performance
pull-request
Mirrored from GitHub Pull Request
 python:uv
refactor
size:L
stale
status: accepting prs
status: blocked
status: in discussion
stt & tts
vulnerablity
web
wont-fix
🌊 feat:workflow
🌚 invalid
🐞 bug
👻 feat:rag
💪 enhancement
📚 documentation
🔨 feat:tools
🙋‍♂️ question
🙏 help wanted
🤔 cant-reproduce
🤖 feat:agent
No Label 🐞 bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langgenius/dify#20350
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?