I have searched the existing issues and discussions.
Problem Description
I am connected my openrouter to the openwebui and I noticed that the model thoughts are not being displayed. I find this infuriating since if the model never send the end-thinking token, I might never see the final outputs. Is there a setting that I am missing?
Desired Solution you'd like
Adding a toggle in the Advanced Parameter setting would be perfect.
Alternatives Considered
No response
Additional Context
No response
Originally created by @LazyDataScientistGit on GitHub (Mar 6, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/11322
### Check Existing Issues
- [x] I have searched the existing issues and discussions.
### Problem Description
I am connected my openrouter to the openwebui and I noticed that the model thoughts are not being displayed. I find this infuriating since if the model never send the end-thinking token, I might never see the final outputs. Is there a setting that I am missing?
### Desired Solution you'd like
Adding a toggle in the Advanced Parameter setting would be perfect.
### Alternatives Considered
_No response_
### Additional Context
_No response_
Do you have Stream Chat Response on? (in settings/models/model-pencil/advanced params)
<!-- gh-comment-id:2705221829 -->
@rgaricano commented on GitHub (Mar 7, 2025):
Do you have _Stream Chat Response_ on? (in settings/models/model-pencil/advanced params)
@LazyDataScientistGit commented on GitHub (Mar 7, 2025):
@rgaricano I have tried both on and off in the settings and none of the reasoning models thoughts are generated when using a model from openrouter, but the model's think does appear when I run the thinking model locally.
<!-- gh-comment-id:2705245308 -->
@LazyDataScientistGit commented on GitHub (Mar 7, 2025):
@rgaricano I have tried both on and off in the settings and none of the reasoning models thoughts are generated when using a model from openrouter, but the model's think does appear when I run the thinking model locally.
If thinking doesn't appaer on local, try indicating on the system prompt or prompt which should show the thinking, thought or reasoning process.
That's how it worked for me.
(Sorry, I misread it, now I see that you said that it works locally)
<!-- gh-comment-id:2705265008 -->
@rgaricano commented on GitHub (Mar 7, 2025):
If thinking doesn't appaer on local, try indicating on the system prompt or prompt which should show the thinking, thought or reasoning process.
That's how it worked for me.
(Sorry, I misread it, now I see that you said that it works locally)
@LazyDataScientistGit commented on GitHub (Mar 7, 2025):
@rgaricano Yeah, I was hoping webui would have the thinking tokens from openrouter, i enjoy reading the thoughts more than the final output.
Here is an example:
<!-- gh-comment-id:2705364022 -->
@LazyDataScientistGit commented on GitHub (Mar 7, 2025):
@rgaricano Yeah, I was hoping webui would have the thinking tokens from openrouter, i enjoy reading the thoughts more than the final output.
Here is an example:
#11259 It is possible that the model did not output the starting label<think>
<!-- gh-comment-id:2705453634 -->
@sleepyddl commented on GitHub (Mar 7, 2025):
#11259 It is possible that the model did not output the starting label`<think>`
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @LazyDataScientistGit on GitHub (Mar 6, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/11322
Check Existing Issues
Problem Description
I am connected my openrouter to the openwebui and I noticed that the model thoughts are not being displayed. I find this infuriating since if the model never send the end-thinking token, I might never see the final outputs. Is there a setting that I am missing?
Desired Solution you'd like
Adding a toggle in the Advanced Parameter setting would be perfect.
Alternatives Considered
No response
Additional Context
No response
@rgaricano commented on GitHub (Mar 7, 2025):
Do you have Stream Chat Response on? (in settings/models/model-pencil/advanced params)
@LazyDataScientistGit commented on GitHub (Mar 7, 2025):
@rgaricano I have tried both on and off in the settings and none of the reasoning models thoughts are generated when using a model from openrouter, but the model's think does appear when I run the thinking model locally.
@rgaricano commented on GitHub (Mar 7, 2025):
If thinking doesn't appaer on local, try indicating on the system prompt or prompt which should show the thinking, thought or reasoning process.
That's how it worked for me.
(Sorry, I misread it, now I see that you said that it works locally)
@LazyDataScientistGit commented on GitHub (Mar 7, 2025):
@rgaricano Yeah, I was hoping webui would have the thinking tokens from openrouter, i enjoy reading the thoughts more than the final output.
Here is an example:
@sleepyddl commented on GitHub (Mar 7, 2025):
#11259 It is possible that the model did not output the starting label
<think>@tjbck commented on GitHub (Mar 7, 2025):
Are you on the latest?