feat: add OpenAI transcribe and TTS models

Signed-off-by: Alexandros Pappas <[email protected]>

Author: apappascs

auto-configurations/models/spring-ai-autoconfigure-model-openai/src/main/java/org/springframework/ai/model/openai/autoconfigure/OpenAiAudioSpeechProperties.java

@@ -36,7 +36,7 @@ public class OpenAiAudioSpeechProperties extends OpenAiParentProperties {
 
 	public static final String CONFIG_PREFIX = "spring.ai.openai.audio.speech";
 
-	public static final String DEFAULT_SPEECH_MODEL = OpenAiAudioApi.TtsModel.TTS_1.getValue();
+	public static final String DEFAULT_SPEECH_MODEL = OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.getValue();
 
 	private static final Float SPEED = 1.0f;
 

auto-configurations/models/spring-ai-autoconfigure-model-openai/src/main/java/org/springframework/ai/model/openai/autoconfigure/OpenAiAudioTranscriptionProperties.java

@@ -26,7 +26,7 @@ public class OpenAiAudioTranscriptionProperties extends OpenAiParentProperties {
 
 	public static final String CONFIG_PREFIX = "spring.ai.openai.audio.transcription";
 
-	public static final String DEFAULT_TRANSCRIPTION_MODEL = OpenAiAudioApi.WhisperModel.WHISPER_1.getValue();
+	public static final String DEFAULT_TRANSCRIPTION_MODEL = OpenAiAudioApi.TranscriptionModels.WHISPER_1.getValue();
 
 	private static final Double DEFAULT_TEMPERATURE = 0.7;
 

models/spring-ai-openai/src/main/java/org/springframework/ai/openai/OpenAiAudioSpeechModel.java

@@ -80,7 +80,7 @@ public class OpenAiAudioSpeechModel implements SpeechModel, StreamingSpeechModel
 	public OpenAiAudioSpeechModel(OpenAiAudioApi audioApi) {
 		this(audioApi,
 				OpenAiAudioSpeechOptions.builder()
-					.model(OpenAiAudioApi.TtsModel.TTS_1.getValue())
+					.model(OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.getValue())
 					.responseFormat(AudioResponseFormat.MP3)
 					.voice(OpenAiAudioApi.SpeechRequest.Voice.ALLOY.getValue())
 					.speed(SPEED)

models/spring-ai-openai/src/main/java/org/springframework/ai/openai/OpenAiAudioTranscriptionModel.java

@@ -63,7 +63,7 @@ public class OpenAiAudioTranscriptionModel implements Model<AudioTranscriptionPr
 	public OpenAiAudioTranscriptionModel(OpenAiAudioApi audioApi) {
 		this(audioApi,
 				OpenAiAudioTranscriptionOptions.builder()
-					.model(OpenAiAudioApi.WhisperModel.WHISPER_1.getValue())
+					.model(OpenAiAudioApi.TranscriptionModels.WHISPER_1.getValue())
 					.responseFormat(OpenAiAudioApi.TranscriptResponseFormat.JSON)
 					.temperature(0.7f)
 					.build());

models/spring-ai-openai/src/main/java/org/springframework/ai/openai/api/OpenAiAudioApi.java

@@ -26,6 +26,7 @@
 import reactor.core.publisher.Mono;
 
 import org.springframework.ai.model.ApiKey;
+import org.springframework.ai.model.ChatModelDescription;
 import org.springframework.ai.model.NoopApiKey;
 import org.springframework.ai.model.SimpleApiKey;
 import org.springframework.ai.openai.api.common.OpenAiApiConstants;
@@ -213,22 +214,25 @@ public String getFilename() {
 	 * different model variates, tts-1 is optimized for real time text to speech use cases
 	 * and tts-1-hd is optimized for quality. These models can be used with the Speech
 	 * endpoint in the Audio API. Reference:
-	 * <a href="https://platform.openai.com/docs/models/tts">TTS</a>
+	 * <a href="https://platform.openai.com/docs/models#tts">TTS</a>
 	 */
 	public enum TtsModel {
 
-		// @formatter:off
 		/**
-		 * The latest text to speech model, optimized for speed.
+		 * Text-to-speech model optimized for speed
 		 */
 		@JsonProperty("tts-1")
 		TTS_1("tts-1"),
 		/**
-		 * The latest text to speech model, optimized for quality.
+		 * Text-to-speech model optimized for quality.
 		 */
 		@JsonProperty("tts-1-hd")
-		TTS_1_HD("tts-1-hd");
-		// @formatter:on
+		TTS_1_HD("tts-1-hd"),
+		/**
+		 * Text-to-speech model powered by GPT-4o mini
+		 */
+		@JsonProperty("gpt-4o-mini-tts")
+		GPT_4O_MINI_TTS("gpt-4o-mini-tts");
 
 		public final String value;
 
@@ -250,6 +254,7 @@ public String getValue() {
 	 * v2-large model is currently available through our API with the whisper-1 model
 	 * name.
 	 */
+	@Deprecated
 	public enum WhisperModel {
 
 		// @formatter:off
@@ -269,6 +274,45 @@ public String getValue() {
 
 	}
 
+	/**
+	 * The available models for the transcriptions API. Reference:
+	 * <a href="https://platform.openai.com/docs/models#transcription">
+	 */
+	public enum TranscriptionModels implements ChatModelDescription {
+
+		/**
+		 * Speech-to-text model powered by GPT-4o
+		 */
+		@JsonProperty("gpt-4o-transcribe")
+		GPT_4O_TRANSCRIBE("gpt-4o-transcribe"),
+		/**
+		 * Speech-to-text model powered by GPT-4o mini
+		 */
+		@JsonProperty("gpt-4o-mini-transcribe")
+		GPT_4O_MINI_TRANSCRIBE("gpt-4o-mini-transcribe"),
+		/**
+		 * General-purpose speech recognition model
+		 */
+		@JsonProperty("whisper-1")
+		WHISPER_1("whisper-1");
+
+		public final String value;
+
+		TranscriptionModels(String value) {
+			this.value = value;
+		}
+
+		public String getValue() {
+			return this.value;
+		}
+
+		@Override
+		public String getName() {
+			return this.value;
+		}
+
+	}
+
 	/**
 	 * The format of the transcript and translation outputs, in one of these options:
 	 * json, text, srt, verbose_json, or vtt. Defaults to json.
@@ -416,7 +460,7 @@ public String getValue() {
 		 */
 		public static class Builder {
 
-			private String model = TtsModel.TTS_1.getValue();
+			private String model = TtsModel.GPT_4O_MINI_TTS.getValue();
 
 			private String input;
 
@@ -531,7 +575,7 @@ public static class Builder {
 
 			private byte[] file;
 
-			private String model = WhisperModel.WHISPER_1.getValue();
+			private String model = TranscriptionModels.WHISPER_1.getValue();
 
 			private String language;
 
@@ -624,7 +668,7 @@ public static class Builder {
 
 			private byte[] file;
 
-			private String model = WhisperModel.WHISPER_1.getValue();
+			private String model = TranscriptionModels.WHISPER_1.getValue();
 
 			private String prompt;
 

models/spring-ai-openai/src/test/java/org/springframework/ai/openai/audio/api/OpenAiAudioApiIT.java

@@ -29,8 +29,8 @@
 import org.springframework.ai.openai.api.OpenAiAudioApi.StructuredResponse;
 import org.springframework.ai.openai.api.OpenAiAudioApi.TranscriptionRequest;
 import org.springframework.ai.openai.api.OpenAiAudioApi.TranslationRequest;
+import org.springframework.ai.openai.api.OpenAiAudioApi.TranscriptionModels;
 import org.springframework.ai.openai.api.OpenAiAudioApi.TtsModel;
-import org.springframework.ai.openai.api.OpenAiAudioApi.WhisperModel;
 import org.springframework.util.FileCopyUtils;
 
 import static org.assertj.core.api.Assertions.assertThat;
@@ -52,7 +52,7 @@ void speechTranscriptionAndTranslation() throws IOException {
 
 		byte[] speech = this.audioApi
 			.createSpeech(SpeechRequest.builder()
-				.model(TtsModel.TTS_1_HD.getValue())
+				.model(TtsModel.GPT_4O_MINI_TTS.getValue())
 				.input("Hello, my name is Chris and I love Spring A.I.")
 				.voice(Voice.ONYX.getValue())
 				.build())
@@ -64,15 +64,15 @@ void speechTranscriptionAndTranslation() throws IOException {
 
 		StructuredResponse translation = this.audioApi
 			.createTranslation(
-					TranslationRequest.builder().model(WhisperModel.WHISPER_1.getValue()).file(speech).build(),
+					TranslationRequest.builder().model(TranscriptionModels.WHISPER_1.getValue()).file(speech).build(),
 					StructuredResponse.class)
 			.getBody();
 
 		assertThat(translation.text().replaceAll(",", "")).isEqualTo("Hello my name is Chris and I love Spring AI.");
 
 		StructuredResponse transcriptionEnglish = this.audioApi
 			.createTranscription(
-					TranscriptionRequest.builder().model(WhisperModel.WHISPER_1.getValue()).file(speech).build(),
+					TranscriptionRequest.builder().model(TranscriptionModels.WHISPER_1.getValue()).file(speech).build(),
 					StructuredResponse.class)
 			.getBody();
 

models/spring-ai-openai/src/test/java/org/springframework/ai/openai/audio/api/OpenAiAudioModelNoOpApiKeysIT.java

@@ -44,7 +44,7 @@ public class OpenAiAudioModelNoOpApiKeysIT {
 	void checkNoOpKey() {
 		assertThatThrownBy(() -> this.audioApi
 			.createSpeech(OpenAiAudioApi.SpeechRequest.builder()
-				.model(OpenAiAudioApi.TtsModel.TTS_1_HD.getValue())
+				.model(OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.getValue())
 				.input("Hello, my name is Chris and I love Spring A.I.")
 				.voice(OpenAiAudioApi.SpeechRequest.Voice.ONYX.getValue())
 				.build())

models/spring-ai-openai/src/test/java/org/springframework/ai/openai/audio/speech/OpenAiSpeechModelIT.java

@@ -64,7 +64,7 @@ void shouldGenerateNonEmptyMp3AudioFromSpeechPrompt() {
 			.voice(OpenAiAudioApi.SpeechRequest.Voice.ALLOY.getValue())
 			.speed(SPEED)
 			.responseFormat(OpenAiAudioApi.SpeechRequest.AudioResponseFormat.MP3)
-			.model(OpenAiAudioApi.TtsModel.TTS_1.value)
+			.model(OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.value)
 			.build();
 		SpeechPrompt speechPrompt = new SpeechPrompt("Today is a wonderful day to build something people love!",
 				speechOptions);
@@ -100,7 +100,7 @@ void speechRateLimitTest() {
 			.voice(OpenAiAudioApi.SpeechRequest.Voice.ALLOY.getValue())
 			.speed(SPEED)
 			.responseFormat(OpenAiAudioApi.SpeechRequest.AudioResponseFormat.MP3)
-			.model(OpenAiAudioApi.TtsModel.TTS_1.value)
+			.model(OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.value)
 			.build();
 		SpeechPrompt speechPrompt = new SpeechPrompt("Today is a wonderful day to build something people love!",
 				speechOptions);
@@ -120,7 +120,7 @@ void shouldStreamNonEmptyResponsesForValidSpeechPrompts() {
 			.voice(OpenAiAudioApi.SpeechRequest.Voice.ALLOY.getValue())
 			.speed(SPEED)
 			.responseFormat(OpenAiAudioApi.SpeechRequest.AudioResponseFormat.MP3)
-			.model(OpenAiAudioApi.TtsModel.TTS_1.value)
+			.model(OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.value)
 			.build();
 
 		SpeechPrompt speechPrompt = new SpeechPrompt("Today is a wonderful day to build something people love!",
@@ -142,7 +142,7 @@ void speechVoicesTest(String voice) {
 			.voice(voice)
 			.speed(SPEED)
 			.responseFormat(OpenAiAudioApi.SpeechRequest.AudioResponseFormat.MP3)
-			.model(OpenAiAudioApi.TtsModel.TTS_1.value)
+			.model(OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.value)
 			.build();
 		SpeechPrompt speechPrompt = new SpeechPrompt("Today is a wonderful day to build something people love!",
 				speechOptions);

models/spring-ai-openai/src/test/java/org/springframework/ai/openai/audio/speech/OpenAiSpeechModelWithSpeechResponseMetadataTests.java

@@ -75,7 +75,7 @@ void aiResponseContainsImageResponseMetadata() {
 			.voice(OpenAiAudioApi.SpeechRequest.Voice.ALLOY.getValue())
 			.speed(SPEED)
 			.responseFormat(OpenAiAudioApi.SpeechRequest.AudioResponseFormat.MP3)
-			.model(OpenAiAudioApi.TtsModel.TTS_1.value)
+			.model(OpenAiAudioApi.TtsModel.GPT_4O_MINI_TTS.value)
 			.build();
 
 		SpeechPrompt speechPrompt = new SpeechPrompt("Today is a wonderful day to build something people love!",