Question about handling workflows with different API keys under the same baseURL #7365
-
|
Hi, Team I’m a beginner and have a question for you. As shown in the screenshot below, I’d like to add several workflows (based on DIFY, COZE, etc.) as backend APIs, where each model corresponds to a separate workflow. The challenge I’m facing is that these workflows share the same baseURL, but each model requires a different API key. How should I handle this scenario? 
I’ve considered a different approach (see the next screenshot), but the problem is that after selecting “Model2,” I still need to pick the model again from a sub-menu. Is there a way to reduce this extra step? 
Any suggestions or best practices would be greatly appreciated. Thank you! |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 4 replies
-
|
Have you tried adding each as a different custom endpoint? |
Beta Was this translation helpful? Give feedback.
-
|
Thank you for your reply, sorry I just saw this message. I've tried it, actually the second picture above is it. But it must set a default model, which leads users to choose a classification and then choose another model, although this model is the only model under the classification. |
Beta Was this translation helpful? Give feedback.
-
|
Look into Model Specs, which will allow you to define single model items on an endpoint basis. https://www.librechat.ai/docs/configuration/librechat_yaml/object_structure/model_specs#overview |
Beta Was this translation helpful? Give feedback.
-
Mode Specs has a limitation, the AI model must be one of the following: openAI, azureOpenAI, google, anthropic, assistants, azureAssistants, bedrock But my model is custom and does not belong to any of the above. |
Beta Was this translation helpful? Give feedback.
-
@danny-avila Hi, Danny Can you please take a look? Thank you for your time! |
Beta Was this translation helpful? Give feedback.
Have you tried adding each as a different custom endpoint?