How to select models "on the fly"?

[ThadeuFerreira]

I've just started with this library and I'm trying to find a way to invoke different models depending on the situation.

Following the tutorials, I have this in my application.properties

spring.ai.openai.embedding.options.model=text-embedding-3-small
spring.ai.openai.chat.options.model=gpt-3.5-turbo-0125
...

And this configuration

    @Bean
    ChatClient chatClient(ChatClient.Builder builder) {
        return builder.build();
    }

I couldn't find a way to change the configurations of ChatClient in Running time.
Should I have multiple Beans for each model?
O can I change it in the Controller?

[abel533]

OpenAiChatOptions options = OpenAiChatOptions.builder().withModel("model-name").build();
Prompt prompt = new Prompt(new UserMessage(message), options);
return Map.of("generation", chatModel.call(prompt));

[ThomasVitale]

Using the ChatClient fluent API:

chatClient
        .prompt()
        .options(OpenAiChatOptions.builder().withModel("gtp-4-turbo").build())
        .user(question)
        .call()
        .content();

[abel533]

I saw the email and thought it was issues pushed by 'cloud-native-spring-in-action'. Upon closer inspection, I realized it was from Spring AI. What a coincidence, 'cloud-native-spring-in-action' was the first book I purchased on Manning.

[ThomasVitale]

@abel533 thanks for sharing that. I hope you found the book useful!