Service Fallback Mechanism

[0xffmeta]

Hi team,

Is there any fallback mechanism in semantic-kernel python lib? I have a use case that requires:

1. call chatgpt o3 mini
2. if step 1 failed/timed out, fallback to chatgpt 4o model

I want to add this fallback logic globally for the semantic kernel in my project - not sure if there is any good way for this.

[sphenry]

Semantic doesn't have built in fallback logic right now, but you can build your own like the sample here: https://github.com/microsoft/semantic-kernel/tree/main/dotnet/samples/Demos/AIModelRouter

[0xffmeta]

Thank you @sphenry - This doc is also very useful https://github.com/microsoft/semantic-kernel/blob/main/docs/decisions/0019-semantic-function-multiple-model-support.md