[Feature] Added per-call headers capabilities to OpenAiClient

pom.xml

@@ -4,7 +4,7 @@
 
  <groupId>dev.ai4j</groupId>
  <artifactId>openai4j</artifactId>
- <version>0.17.0</version>
+ <version>0.18.0</version>
 
  <name>Java Client for OpenAI (ChatGPT)</name>
  <description>Java Client for OpenAI (ChatGPT)</description>

src/main/java/dev/ai4j/openai4j/DefaultOpenAiClient.java

@@ -128,11 +128,12 @@ public DefaultOpenAiClient build() {
  }
 
  @Override
- public SyncOrAsyncOrStreaming<CompletionResponse> completion(CompletionRequest request) {
+ public SyncOrAsyncOrStreaming<CompletionResponse> completion(OpenAiClientContext context,
+ CompletionRequest request) {
  CompletionRequest syncRequest = CompletionRequest.builder().from(request).stream(null).build();
 
  return new RequestExecutor<>(
- openAiApi.completions(syncRequest, apiVersion),
+ openAiApi.completions(context.headers(), syncRequest, apiVersion),
  r -> r,
  okHttpClient,
  formatUrl("completions"),
@@ -144,13 +145,13 @@ public SyncOrAsyncOrStreaming<CompletionResponse> completion(CompletionRequest r
  }
 
  @Override
- public SyncOrAsyncOrStreaming<String> completion(String prompt) {
+ public SyncOrAsyncOrStreaming<String> completion(OpenAiClientContext context, String prompt) {
  CompletionRequest request = CompletionRequest.builder().prompt(prompt).build();
 
  CompletionRequest syncRequest = CompletionRequest.builder().from(request).stream(null).build();
 
  return new RequestExecutor<>(
- openAiApi.completions(syncRequest, apiVersion),
+ openAiApi.completions(context.headers(), syncRequest, apiVersion),
  CompletionResponse::text,
  okHttpClient,
  formatUrl("completions"),
@@ -162,11 +163,12 @@ public SyncOrAsyncOrStreaming<String> completion(String prompt) {
  }
 
  @Override
- public SyncOrAsyncOrStreaming<ChatCompletionResponse> chatCompletion(ChatCompletionRequest request) {
+ public SyncOrAsyncOrStreaming<ChatCompletionResponse> chatCompletion(OpenAiClientContext context,
+ ChatCompletionRequest request) {
  ChatCompletionRequest syncRequest = ChatCompletionRequest.builder().from(request).stream(null).build();
 
  return new RequestExecutor<>(
- openAiApi.chatCompletions(syncRequest, apiVersion),
+ openAiApi.chatCompletions(context.headers(), syncRequest, apiVersion),
  r -> r,
  okHttpClient,
  formatUrl("chat/completions"),
@@ -178,13 +180,13 @@ public SyncOrAsyncOrStreaming<ChatCompletionResponse> chatCompletion(ChatComplet
  }
 
  @Override
- public SyncOrAsyncOrStreaming<String> chatCompletion(String userMessage) {
+ public SyncOrAsyncOrStreaming<String> chatCompletion(OpenAiClientContext context, String userMessage) {
  ChatCompletionRequest request = ChatCompletionRequest.builder().addUserMessage(userMessage).build();
 
  ChatCompletionRequest syncRequest = ChatCompletionRequest.builder().from(request).stream(null).build();
 
  return new RequestExecutor<>(
- openAiApi.chatCompletions(syncRequest, apiVersion),
+ openAiApi.chatCompletions(context.headers(), syncRequest, apiVersion),
  ChatCompletionResponse::content,
  okHttpClient,
  formatUrl("chat/completions"),
@@ -196,32 +198,38 @@ public SyncOrAsyncOrStreaming<String> chatCompletion(String userMessage) {
  }
 
  @Override
- public SyncOrAsync<EmbeddingResponse> embedding(EmbeddingRequest request) {
- return new RequestExecutor<>(openAiApi.embeddings(request, apiVersion), r -> r);
+ public SyncOrAsync<EmbeddingResponse> embedding(OpenAiClientContext context, EmbeddingRequest request) {
+ return new RequestExecutor<>(openAiApi.embeddings(context.headers(), request, apiVersion), r -> r);
  }
 
  @Override
- public SyncOrAsync<List<Float>> embedding(String input) {
+ public SyncOrAsync<List<Float>> embedding(OpenAiClientContext context, String input) {
  EmbeddingRequest request = EmbeddingRequest.builder().input(input).build();
 
- return new RequestExecutor<>(openAiApi.embeddings(request, apiVersion), EmbeddingResponse::embedding);
+ return new RequestExecutor<>(openAiApi.embeddings(context.headers(), request, apiVersion),
+ EmbeddingResponse::embedding);
  }
 
  @Override
- public SyncOrAsync<ModerationResponse> moderation(ModerationRequest request) {
- return new RequestExecutor<>(openAiApi.moderations(request, apiVersion), r -> r);
+ public SyncOrAsync<ModerationResponse> moderation(OpenAiClientContext context,
+ ModerationRequest request) {
+ return new RequestExecutor<>(openAiApi.moderations(context.headers(), request, apiVersion),
+ r -> r);
  }
 
  @Override
- public SyncOrAsync<ModerationResult> moderation(String input) {
+ public SyncOrAsync<ModerationResult> moderation(OpenAiClientContext context, String input) {
  ModerationRequest request = ModerationRequest.builder().input(input).build();
 
- return new RequestExecutor<>(openAiApi.moderations(request, apiVersion), r -> r.results().get(0));
+ return new RequestExecutor<>(openAiApi.moderations(context.headers(), request, apiVersion),
+ r -> r.results().get(0));
  }
 
  @Override
- public SyncOrAsync<GenerateImagesResponse> imagesGeneration(GenerateImagesRequest request) {
- return new RequestExecutor<>(openAiApi.imagesGenerations(request, apiVersion), r -> r);
+ public SyncOrAsync<GenerateImagesResponse> imagesGeneration(OpenAiClientContext context,
+ GenerateImagesRequest request) {
+ return new RequestExecutor<>(openAiApi.imagesGenerations(context.headers(), request, apiVersion),
+ r -> r);
  }
 
  private String formatUrl(String endpoint) {

src/main/java/dev/ai4j/openai4j/OpenAiApi.java

@@ -10,16 +10,26 @@
 import dev.ai4j.openai4j.image.GenerateImagesResponse;
 import dev.ai4j.openai4j.moderation.ModerationRequest;
 import dev.ai4j.openai4j.moderation.ModerationResponse;
+import java.util.Map;
 import retrofit2.Call;
 import retrofit2.http.Body;
+import retrofit2.http.HeaderMap;
 import retrofit2.http.Headers;
 import retrofit2.http.POST;
 import retrofit2.http.Query;
 
 interface OpenAiApi {
  @POST("completions")
  @Headers("Content-Type: application/json")
- Call<CompletionResponse> completions(@Body CompletionRequest request, @Query("api-version") String apiVersion);
+ Call<CompletionResponse> completions(@Body CompletionRequest request,
+ @Query("api-version") String apiVersion);
+
+ @POST("completions")
+ @Headers("Content-Type: application/json")
+ Call<CompletionResponse> completions(
+ @HeaderMap Map<String, String> headers,
+ @Body CompletionRequest request,
+ @Query("api-version") String apiVersion);
 
  @POST("chat/completions")
  @Headers("Content-Type: application/json")
@@ -28,17 +38,51 @@ Call<ChatCompletionResponse> chatCompletions(
  @Query("api-version") String apiVersion
  );
 
+ @POST("chat/completions")
+ @Headers("Content-Type: application/json")
+ Call<ChatCompletionResponse> chatCompletions(
+ @HeaderMap Map<String, String> headers,
+ @Body ChatCompletionRequest request,
+ @Query("api-version") String apiVersion
+ );
+
+ @POST("embeddings")
+ @Headers("Content-Type: application/json")
+ Call<EmbeddingResponse> embeddings(
+ @Body EmbeddingRequest request,
+ @Query("api-version") String apiVersion);
+
  @POST("embeddings")
  @Headers("Content-Type: application/json")
- Call<EmbeddingResponse> embeddings(@Body EmbeddingRequest request, @Query("api-version") String apiVersion);
+ Call<EmbeddingResponse> embeddings(
+ @HeaderMap Map<String, String> headers,
+ @Body EmbeddingRequest request,
+ @Query("api-version") String apiVersion);
+
+ @POST("moderations")
+ @Headers("Content-Type: application/json")
+ Call<ModerationResponse> moderations(
+ @Body ModerationRequest request,
+ @Query("api-version") String apiVersion);
 
  @POST("moderations")
  @Headers("Content-Type: application/json")
- Call<ModerationResponse> moderations(@Body ModerationRequest request, @Query("api-version") String apiVersion);
+ Call<ModerationResponse> moderations(
+ @HeaderMap Map<String, String> headers,
+ @Body ModerationRequest request,
+ @Query("api-version") String apiVersion);
+
+ @POST("images/generations")
+ @Headers({"Content-Type: application/json"})
+ Call<GenerateImagesResponse> imagesGenerations(
+ @Body GenerateImagesRequest request,
+ @Query("api-version") String apiVersion
+ );
 
  @POST("images/generations")
- @Headers({ "Content-Type: application/json" })
+ @Headers({"Content-Type: application/json"})
  Call<GenerateImagesResponse> imagesGenerations(
+ @HeaderMap Map<String, String> headers,
  @Body GenerateImagesRequest request,
  @Query("api-version") String apiVersion
  );

src/main/java/dev/ai4j/openai4j/OpenAiClient.java

@@ -1,11 +1,6 @@
 package dev.ai4j.openai4j;
 
-import java.net.InetSocketAddress;
-import java.net.Proxy;
-import java.nio.file.Path;
-import java.nio.file.Paths;
-import java.time.Duration;
-import java.util.List;
+import static dev.ai4j.openai4j.LogLevel.DEBUG;
 
 import dev.ai4j.openai4j.chat.ChatCompletionRequest;
 import dev.ai4j.openai4j.chat.ChatCompletionResponse;
@@ -20,29 +15,101 @@
 import dev.ai4j.openai4j.moderation.ModerationResult;
 import dev.ai4j.openai4j.spi.OpenAiClientBuilderFactory;
 import dev.ai4j.openai4j.spi.ServiceHelper;
+import java.net.InetSocketAddress;
+import java.net.Proxy;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.time.Duration;
+import java.util.HashMap;
+import java.util.List;
 import java.util.Map;
 
-import static dev.ai4j.openai4j.LogLevel.DEBUG;
-
 public abstract class OpenAiClient {
 
- public abstract SyncOrAsyncOrStreaming<CompletionResponse> completion(CompletionRequest request);
+ public SyncOrAsyncOrStreaming<CompletionResponse> completion(CompletionRequest request) {
+ return completion(new OpenAiClientContext(), request);
+ }
+
+ public SyncOrAsyncOrStreaming<CompletionResponse> completion(
+ OpenAiClientContext clientContext, CompletionRequest request) {
+ throw new UnsupportedOperationException();
+ }
+
+ public SyncOrAsyncOrStreaming<String> completion(String prompt) {
+ return completion(new OpenAiClientContext(), prompt);
+ }
+
+ public SyncOrAsyncOrStreaming<String> completion(OpenAiClientContext clientContext,
+ String prompt) {
+ throw new UnsupportedOperationException();
+ }
+
+ public SyncOrAsyncOrStreaming<ChatCompletionResponse> chatCompletion(
+ ChatCompletionRequest request) {
+ return chatCompletion(new OpenAiClientContext(), request);
+ }
+
+ public SyncOrAsyncOrStreaming<ChatCompletionResponse> chatCompletion(
+ OpenAiClientContext clientContext,
+ ChatCompletionRequest request) {
+ throw new UnsupportedOperationException();
+ }
+
+ public SyncOrAsyncOrStreaming<String> chatCompletion(String userMessage) {
+ return chatCompletion(new OpenAiClientContext(), userMessage);
+ }
 
- public abstract SyncOrAsyncOrStreaming<String> completion(String prompt);
+ public SyncOrAsyncOrStreaming<String> chatCompletion(
+ OpenAiClientContext clientContext,
+ String userMessage) {
+ throw new UnsupportedOperationException();
+ }
 
- public abstract SyncOrAsyncOrStreaming<ChatCompletionResponse> chatCompletion(ChatCompletionRequest request);
+ public SyncOrAsync<EmbeddingResponse> embedding(EmbeddingRequest request) {
+ return embedding(new OpenAiClientContext(), request);
+ }
 
- public abstract SyncOrAsyncOrStreaming<String> chatCompletion(String userMessage);
+ public SyncOrAsync<EmbeddingResponse> embedding(OpenAiClientContext clientContext,
+ EmbeddingRequest request) {
+ throw new UnsupportedOperationException();
+ }
 
- public abstract SyncOrAsync<EmbeddingResponse> embedding(EmbeddingRequest request);
+ public SyncOrAsync<List<Float>> embedding(String input) {
+ return embedding(new OpenAiClientContext(), input);
+ }
 
- public abstract SyncOrAsync<List<Float>> embedding(String input);
+ public SyncOrAsync<List<Float>> embedding(OpenAiClientContext clientContext,
+ String input) {
+ throw new UnsupportedOperationException();
+ }
+
+ public SyncOrAsync<ModerationResponse> moderation(ModerationRequest request) {
+ return moderation(new OpenAiClientContext(), request);
+ }
+
+ public SyncOrAsync<ModerationResponse> moderation(OpenAiClientContext clientContext,
+ ModerationRequest request) {
+ throw new UnsupportedOperationException();
+ }
+
+ public SyncOrAsync<ModerationResult> moderation(String input) {
+ return moderation(new OpenAiClientContext(), input);
+ }
 
- public abstract SyncOrAsync<ModerationResponse> moderation(ModerationRequest request);
+ public SyncOrAsync<ModerationResult> moderation(OpenAiClientContext clientContext,
+ String input) {
+ throw new UnsupportedOperationException();
+ }
 
- public abstract SyncOrAsync<ModerationResult> moderation(String input);
+ public SyncOrAsync<GenerateImagesResponse> imagesGeneration(GenerateImagesRequest request) {
+ return imagesGeneration(new OpenAiClientContext(), request);
+ }
 
- public abstract SyncOrAsync<GenerateImagesResponse> imagesGeneration(GenerateImagesRequest request);
+ public SyncOrAsync<GenerateImagesResponse> imagesGeneration(
+ OpenAiClientContext clientContext,
+ GenerateImagesRequest request) {
+ throw new UnsupportedOperationException();
+ }
 
  public abstract void shutdown();
 
@@ -55,6 +122,28 @@ public static OpenAiClient.Builder builder() {
  return DefaultOpenAiClient.builder();
  }
 
+ public static class OpenAiClientContext {
+ private final Map<String, String> headers = new HashMap<>();
+
+ public OpenAiClientContext addHeaders(Map<String, String> headers) {
+ this.headers.putAll(headers);
+ return this;
+ }
+
+ public OpenAiClientContext addHeader(String key, String value) {
+ headers.put(key, value);
+ return this;
+ }
+
+ public Map<String, String> headers() {
+ return headers;
+ }
+
+ public static OpenAiClientContext create() {
+ return new OpenAiClientContext();
+ }
+ }
+
  @SuppressWarnings("unchecked")
  public abstract static class Builder<T extends OpenAiClient, B extends Builder<T, B>> {
 

src/main/java/dev/ai4j/openai4j/chat/ChatCompletionModel.java

@@ -3,17 +3,15 @@
 public enum ChatCompletionModel {
 
  GPT_3_5_TURBO("gpt-3.5-turbo"), // alias
- GPT_3_5_TURBO_0613("gpt-3.5-turbo-0613"),
  GPT_3_5_TURBO_1106("gpt-3.5-turbo-1106"),
  GPT_3_5_TURBO_0125("gpt-3.5-turbo-0125"),
 
- GPT_3_5_TURBO_16K("gpt-3.5-turbo-16k"), // alias
- GPT_3_5_TURBO_16K_0613("gpt-3.5-turbo-16k-0613"),
-
  GPT_4("gpt-4"), // alias
  GPT_4_0314("gpt-4-0314"),
  GPT_4_0613("gpt-4-0613"),
 
+ GPT_4_TURBO("gpt-4-turbo"), // alias
+ GPT_4_TURBO_2024_04_09("gpt-4-turbo-2024-04-09"), // With vision support
  GPT_4_TURBO_PREVIEW("gpt-4-turbo-preview"), // alias
  GPT_4_1106_PREVIEW("gpt-4-1106-preview"),
  GPT_4_0125_PREVIEW("gpt-4-0125-preview"),
@@ -22,7 +20,10 @@ public enum ChatCompletionModel {
  GPT_4_32K_0314("gpt-4-32k-0314"),
  GPT_4_32K_0613("gpt-4-32k-0613"),
 
- GPT_4_VISION_PREVIEW("gpt-4-vision-preview");
+ @Deprecated
+ GPT_4_VISION_PREVIEW("gpt-4-vision-preview"),
+ GPT_4O("gpt-4o"),
+ GPT_4O_2024_05_13("gpt-4o-2024-05-13");
 
  private final String value;