I have searched the existing issues and discussions.
Problem Description
When connecting multiple local Ollama servers in Open WebUI, it’s possible for the same model (e.g., llama3) to exist on more than one server. Currently, there’s no way to:
See that the same model exists on multiple servers
Choose which server should be used when starting a chat or inference
As a result, Open WebUI defaults to one of the servers arbitrarily, making it hard to control or understand where the workload is being run.
My current workaround is to disable the second server so I can guarantee that it will use the one that I desire.
Desired Solution you'd like
It would be helpful if Open WebUI:
Displays duplicate models in the model selector
Labels them clearly (e.g., by IP address, hostname, or a custom server label)
Allows the user to explicitly choose which server to run the model from
This might be as simple as changing the model dropdown UI to group duplicate model names by server.
Alternatives Considered
A “preferred server” setting in the model configuration (but less flexible for switching)
Additional Context
No response
Originally created by @TigerWolf on GitHub (Jul 16, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/15755
### Check Existing Issues
- [x] I have searched the existing issues and discussions.
### Problem Description
When connecting multiple local Ollama servers in Open WebUI, it’s possible for the same model (e.g., llama3) to exist on more than one server. Currently, there’s no way to:
- See that the same model exists on multiple servers
- Choose which server should be used when starting a chat or inference
As a result, Open WebUI defaults to one of the servers arbitrarily, making it hard to control or understand where the workload is being run.
My current workaround is to disable the second server so I can guarantee that it will use the one that I desire.
### Desired Solution you'd like
It would be helpful if Open WebUI:
- Displays duplicate models in the model selector
- Labels them clearly (e.g., by IP address, hostname, or a custom server label)
- Allows the user to explicitly choose which server to run the model from
This might be as simple as changing the model dropdown UI to group duplicate model names by server.
### Alternatives Considered
- A “preferred server” setting in the model configuration (but less flexible for switching)
### Additional Context
_No response_
There are an option for prefix connections to distinguish them, in connection settings.
<!-- gh-comment-id:3076544385 -->
@rgaricano commented on GitHub (Jul 16, 2025):
There are an option for prefix connections to distinguish them, in connection settings.
It would be nice to make it easier to see which one without having to hover over them both. Thoughts @rgaricano ?
<!-- gh-comment-id:3076564970 -->
@TigerWolf commented on GitHub (Jul 16, 2025):
Fantastic - this solves my problem.
It would be nice to make it easier to see which one without having to hover over them both. Thoughts @rgaricano ?
<img width="541" height="101" alt="Image" src="https://github.com/user-attachments/assets/764bdc0f-322b-4df0-bb09-431574cdbac8" />
<!-- gh-comment-id:3076573669 -->
@rgaricano commented on GitHub (Jul 16, 2025):
Yes, I think so, I did a PR https://github.com/open-webui/open-webui/pull/12860 so that the prefix was more visible in list but was rejected, right now that is the default behaviour.
I will close this for now. I had a quick look at the code and I dont think there are any great ways to implement this without making the UI too messy.
<!-- gh-comment-id:3076591163 -->
@TigerWolf commented on GitHub (Jul 16, 2025):
I will close this for now. I had a quick look at the code and I dont think there are any great ways to implement this without making the UI too messy.
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 @TigerWolf on GitHub (Jul 16, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/15755
Check Existing Issues
Problem Description
When connecting multiple local Ollama servers in Open WebUI, it’s possible for the same model (e.g., llama3) to exist on more than one server. Currently, there’s no way to:
As a result, Open WebUI defaults to one of the servers arbitrarily, making it hard to control or understand where the workload is being run.
My current workaround is to disable the second server so I can guarantee that it will use the one that I desire.
Desired Solution you'd like
It would be helpful if Open WebUI:
This might be as simple as changing the model dropdown UI to group duplicate model names by server.
Alternatives Considered
Additional Context
No response
@rgaricano commented on GitHub (Jul 16, 2025):
There are an option for prefix connections to distinguish them, in connection settings.
@TigerWolf commented on GitHub (Jul 16, 2025):
Fantastic - this solves my problem.
It would be nice to make it easier to see which one without having to hover over them both. Thoughts @rgaricano ?
@rgaricano commented on GitHub (Jul 16, 2025):
Yes, I think so, I did a PR https://github.com/open-webui/open-webui/pull/12860 so that the prefix was more visible in list but was rejected, right now that is the default behaviour.
@TigerWolf commented on GitHub (Jul 16, 2025):
I will close this for now. I had a quick look at the code and I dont think there are any great ways to implement this without making the UI too messy.