Is it possible to use the auto-configure feature to configure several AI clients from the same provider but with different models, e.g. Anthropic + claude-instant-v1 and Anthropic + anthropic.claude-v2 ?
So far, only the possibility of defining one such combination with auto-configure has been described on the docs pages.
Of course, this could be solved via the low-level API with programmatic creation of the multiple corresponding BedrockApi client. But then the auto-configure feature would not be usable out-of-the-box.
On the other hand, OpenAI offers the option of specifying the model to be used as a chat request option with OpenAiChatOptions#model. If something like this were also possible for AWS Bedrock, then you wouldn't even need the multiple configurations option.
I checked the API and documentation, but didn't find a solution.