<ul><li>Semantic Kernel supports registering multiple text generation and embedding services using serviceId and modelId.</li><li>When executing kernel functions or prompts, 'serviceId' and 'modelId' can be passed into 'PromptExecutionSettings'.</li><li>When registering chat completion services, if serviceId is provided, Semantic Kernel also registers chat completion services as keyed.</li><li>All AI-related services implement the IAIService interface, and custom IAIServiceSelector can be implemented for full control over service selection.</li></ul>

Working with multiple language models in Semantic Kernel

Discover more