Package dev.agenor.adapters.llm.anthropic

Class AnthropicProvider

java.lang.Object

dev.agenor.adapters.llm.anthropic.AnthropicProvider

All Implemented Interfaces:

LLMProvider

public class AnthropicProvider
extends Object
implements LLMProvider

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	AnthropicProvider.Builder
static enum	AnthropicProvider.Models	Deprecated, for removal: This API element is subject to removal in a future version. since 0.36.0, for removal in 1.0.0.

Method Summary

Modifier and Type	Method	Description
static AnthropicProvider.Builder	builder()
CompletableFuture<LLMResponse>	chat(LLMRequest request)	Send a chat completion request to the LLM.
CompletableFuture<Void>	chatStream(LLMRequest request, Consumer<StreamingChunk> handler)	Stream a chat completion request to the LLM.
CompletableFuture<List<String>>	getAvailableModels()	Return the models this provider recognises.
String	getDefaultModel()	Get the default model for this provider.
String	getProviderName()	Get the name of this provider.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface dev.agenor.core.llm.LLMProvider

supportsFunctionCalling, supportsStreaming, validateRequest

Method Details

chat

public CompletableFuture<LLMResponse> chat(LLMRequest request)

Description copied from interface: LLMProvider

Send a chat completion request to the LLM.

This is the primary method for interacting with the LLM. It sends a request containing messages and configuration, and returns a response asynchronously.

Specified by:

chat in interface LLMProvider

Returns:

CompletableFuture containing the LLM response

chatStream

public CompletableFuture<Void> chatStream(LLMRequest request,
 Consumer<StreamingChunk> handler)

Description copied from interface: LLMProvider

Stream a chat completion request to the LLM.

Similar to LLMProvider.chat(LLMRequest), but provides streaming updates as the response is generated. This is useful for long responses where you want to show incremental progress to users.

Example usage:

 provider.chatStream(request, chunk -> {
     System.out.print(chunk.content());
 }).thenRun(() -> {
     System.out.println("\nStreaming complete");
 });
 

Specified by:

chatStream in interface LLMProvider

Returns:

CompletableFuture that completes when streaming is finished

getAvailableModels

public CompletableFuture<List<String>> getAvailableModels()

Description copied from interface: LLMProvider

Return the models this provider recognises.

Model names are provider-specific (e.g., "gpt-4", "claude-3-opus", "llama2"). This is not a live query against the provider's API: an adapter may report a fixed list it ships with, so a model missing from it can still be requested by name — see LLMRequest.Builder.model(String) — it simply will not appear here.

Specified by:

getAvailableModels in interface LLMProvider

Returns:

CompletableFuture containing the model identifiers this provider recognises

getProviderName

public String getProviderName()

Description copied from interface: LLMProvider

Get the name of this provider.

Returns a human-readable identifier for this provider, such as "OpenAI", "Anthropic", or "Ollama".

Specified by:

getProviderName in interface LLMProvider

Returns:

the provider name

getDefaultModel

public String getDefaultModel()

Description copied from interface: LLMProvider

Get the default model for this provider.

Returns a reasonable default model that can be used if the user doesn't specify one. This should be a balanced model that works well for general purposes.

Specified by:

getDefaultModel in interface LLMProvider

Returns:

the default model name, or null if no default is available

builder

public static AnthropicProvider.Builder builder()