Feat: Model Check by Name not ID. #4806

New Issue

2025-11-11T16:03:31-06:00

GiteaMirror commented

2025-11-11 16:03:31 -06:00

Originally created by @spammenotinoz on GitHub (Apr 11, 2025).

Check Existing Issues

I have searched the existing issues and discussions.

Problem Description

I know this can be implemented outside of Open-WebUI and I caused this issue myself. But I do like to experiment with pipelines\providers.

Background:
Model Defined as gpt-4o using provider
I want to switch providers, say using a pipeline and create a new model called gpt-4o.
When I delete the existing model gpt-4o, as expected existing chats using gpt-4o return "model not found" as the model ID changed.
The issue does not occur when the models are both provided by Connections, however if the second is provided by a pipe \ pipeline the failover does not occur.

Workaround:
Not a major issue, as you can just select a new model again.

Desired Solution you'd like

Requested Change:
Instead of throwing the model not found error, check for an existing model of the same name and if found. ie: Model Name.

Similar to how the loadbalancing\failover works for connections.

Alternatives Considered

No response

Additional Context

No response

Originally created by @spammenotinoz on GitHub (Apr 11, 2025). ### Check Existing Issues - [x] I have searched the existing issues and discussions. ### Problem Description I know this can be implemented outside of Open-WebUI and I caused this issue myself. But I do like to experiment with pipelines\providers. Background: Model Defined as gpt-4o using provider <openai> I want to switch providers, say using a pipeline and create a new model called gpt-4o. When I delete the existing model gpt-4o, as expected existing chats using gpt-4o return "model not found" as the model ID changed. The issue does not occur when the models are both provided by Connections, however if the second is provided by a pipe \ pipeline the failover does not occur. Workaround: Not a major issue, as you can just select a new model again. ### Desired Solution you'd like Requested Change: Instead of throwing the model not found error, check for an existing model of the same name and if found. ie: Model Name. Similar to how the loadbalancing\failover works for connections. ### Alternatives Considered _No response_ ### Additional Context _No response_

GiteaMirror closed this issue

2025-11-11 16:03:32 -06:00

GiteaMirror referenced this issue

2026-04-19 20:22:20 -05:00

[GH-ISSUE #4806] Error with some embedding models with RAG (files not found) #13738

GiteaMirror referenced this issue

2026-04-25 03:41:35 -05:00

[GH-ISSUE #4806] Error with some embedding models with RAG (files not found) #29266

GiteaMirror referenced this issue

2026-05-05 13:29:35 -05:00

[GH-ISSUE #4806] Error with some embedding models with RAG (files not found) #52404

GiteaMirror referenced this issue