Originally created by @zhiqing0205 on GitHub (Aug 15, 2024).
Is your feature request related to a problem? Please describe.
I'm frustrated because some API providers do not support the /v1/models endpoint to retrieve available models, but they do support the /v1/chat/completions endpoint for normal chat interactions. This limitation prevents the use of certain API providers with OpenWebUI, as it currently relies on the /v1/models endpoint to detect available models.
Describe the solution you'd like
I would like to request the ability to manually customize or input a model in the top left corner of the interface. This feature would allow users to specify the model they want to use when the API provider does not support the /v1/models endpoint, enabling broader compatibility with different API providers.
Describe alternatives you've considered
An alternative could be implementing a fallback mechanism where, if the /v1/models endpoint is not supported by the API provider, users could still input a model manually or select from a predefined list of commonly used models.
This enhancement would greatly improve flexibility and compatibility with a wider range of API providers.
Originally created by @zhiqing0205 on GitHub (Aug 15, 2024).
**Is your feature request related to a problem? Please describe.**
I'm frustrated because some API providers do not support the `/v1/models` endpoint to retrieve available models, but they do support the `/v1/chat/completions` endpoint for normal chat interactions. This limitation prevents the use of certain API providers with OpenWebUI, as it currently relies on the `/v1/models` endpoint to detect available models.
**Describe the solution you'd like**
I would like to request the ability to manually customize or input a model in the top left corner of the interface. This feature would allow users to specify the model they want to use when the API provider does not support the `/v1/models` endpoint, enabling broader compatibility with different API providers.
**Describe alternatives you've considered**
An alternative could be implementing a fallback mechanism where, if the `/v1/models` endpoint is not supported by the API provider, users could still input a model manually or select from a predefined list of commonly used models.
This enhancement would greatly improve flexibility and compatibility with a wider range of API providers.
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 @zhiqing0205 on GitHub (Aug 15, 2024).
Is your feature request related to a problem? Please describe.
I'm frustrated because some API providers do not support the
/v1/modelsendpoint to retrieve available models, but they do support the/v1/chat/completionsendpoint for normal chat interactions. This limitation prevents the use of certain API providers with OpenWebUI, as it currently relies on the/v1/modelsendpoint to detect available models.Describe the solution you'd like
I would like to request the ability to manually customize or input a model in the top left corner of the interface. This feature would allow users to specify the model they want to use when the API provider does not support the
/v1/modelsendpoint, enabling broader compatibility with different API providers.Describe alternatives you've considered
An alternative could be implementing a fallback mechanism where, if the
/v1/modelsendpoint is not supported by the API provider, users could still input a model manually or select from a predefined list of commonly used models.This enhancement would greatly improve flexibility and compatibility with a wider range of API providers.
@moblangeois commented on GitHub (Aug 15, 2024):
In addition, I'd like to point out that it is the case with perplexity API.