@mastra/voice-google 0.0.0-toolOptionTypes-20250917085558 → 0.0.0-top-level-fix-20251211111608

package/dist/index.cjs

@@ -6,61 +6,129 @@ var textToSpeech = require('@google-cloud/text-to-speech');
 var voice = require('@mastra/core/voice');
 
 // src/index.ts
+var resolveAuthConfig = (modelConfig, fallback, vertexConfig) => {
+  const resolved = {};
+  if (vertexConfig?.vertexAI) {
+    const projectId = vertexConfig.project || process.env.GOOGLE_CLOUD_PROJECT;
+    if (projectId) {
+      resolved.projectId = projectId;
+    }
+  }
+  const apiKey = modelConfig?.apiKey ?? fallback.apiKey;
+  if (apiKey && !vertexConfig?.vertexAI) {
+    resolved.apiKey = apiKey;
+  }
+  const keyFilename = modelConfig?.keyFilename ?? fallback.keyFilename;
+  if (keyFilename) {
+    resolved.keyFilename = keyFilename;
+  }
+  const credentials = modelConfig?.credentials ?? fallback.credentials;
+  if (credentials) {
+    resolved.credentials = credentials;
+  }
+  return resolved;
+};
+var buildAuthOptions = (config, vertexConfig) => {
+  const options = {};
+  if (config.credentials) {
+    options.credentials = config.credentials;
+  }
+  if (config.keyFilename) {
+    options.keyFilename = config.keyFilename;
+  }
+  if (config.apiKey && !vertexConfig?.vertexAI) {
+    options.apiKey = config.apiKey;
+  }
+  if (config.projectId) {
+    options.projectId = config.projectId;
+  }
+  return options;
+};
 var DEFAULT_VOICE = "en-US-Casual-K";
 var GoogleVoice = class extends voice.MastraVoice {
   ttsClient;
   speechClient;
+  vertexAI;
+  project;
+  location;
   /**
    * Creates an instance of GoogleVoice
-   * @param {Object} config - Configuration options
+   * @param {GoogleVoiceConfig} config - Configuration options
    * @param {GoogleModelConfig} [config.speechModel] - Configuration for speech synthesis
    * @param {GoogleModelConfig} [config.listeningModel] - Configuration for speech recognition
    * @param {string} [config.speaker] - Default voice ID to use for speech synthesis
-   * @throws {Error} If no API key is provided via config or environment variable
+   * @param {boolean} [config.vertexAI] - Enable Vertex AI mode
+   * @param {string} [config.project] - Google Cloud project ID (required for Vertex AI)
+   * @param {string} [config.location] - Google Cloud region (default: 'us-central1')
    */
-  constructor({
-    listeningModel,
-    speechModel,
-    speaker
-  } = {}) {
+  constructor({ listeningModel, speechModel, speaker, vertexAI = false, project, location } = {}) {
     const defaultApiKey = process.env.GOOGLE_API_KEY;
+    const defaultKeyFilename = process.env.GOOGLE_APPLICATION_CREDENTIALS;
     const defaultSpeaker = DEFAULT_VOICE;
+    const resolvedProject = project || process.env.GOOGLE_CLOUD_PROJECT;
+    const resolvedLocation = location || process.env.GOOGLE_CLOUD_LOCATION || "us-central1";
+    if (vertexAI && !resolvedProject) {
+      throw new Error(
+        "Google Cloud project ID is required when using Vertex AI. Set GOOGLE_CLOUD_PROJECT environment variable or pass project to constructor."
+      );
+    }
+    const vertexConfig = { vertexAI, project: resolvedProject };
+    const sharedFallback = {
+      apiKey: defaultApiKey ?? speechModel?.apiKey ?? listeningModel?.apiKey,
+      keyFilename: defaultKeyFilename ?? speechModel?.keyFilename ?? listeningModel?.keyFilename,
+      credentials: speechModel?.credentials ?? listeningModel?.credentials};
+    const speechAuthConfig = resolveAuthConfig(speechModel, sharedFallback, vertexConfig);
+    const listeningAuthConfig = resolveAuthConfig(listeningModel, sharedFallback, vertexConfig);
     super({
       speechModel: {
         name: "",
-        apiKey: speechModel?.apiKey ?? defaultApiKey
+        apiKey: speechAuthConfig.apiKey ?? defaultApiKey
       },
       listeningModel: {
         name: "",
-        apiKey: listeningModel?.apiKey ?? defaultApiKey
+        apiKey: listeningAuthConfig.apiKey ?? defaultApiKey
       },
       speaker: speaker ?? defaultSpeaker
     });
-    const apiKey = defaultApiKey || speechModel?.apiKey || listeningModel?.apiKey;
-    if (!apiKey) {
-      throw new Error(
-        "Google API key is not set, set GOOGLE_API_KEY environment variable or pass apiKey to constructor"
-      );
-    }
-    this.ttsClient = new textToSpeech.TextToSpeechClient({
-      apiKey: this.speechModel?.apiKey || defaultApiKey
-    });
-    this.speechClient = new speech.SpeechClient({
-      apiKey: this.listeningModel?.apiKey || defaultApiKey
-    });
+    this.vertexAI = vertexAI;
+    this.project = resolvedProject;
+    this.location = resolvedLocation;
+    const ttsOptions = buildAuthOptions(speechAuthConfig, { vertexAI});
+    const speechOptions = buildAuthOptions(listeningAuthConfig, { vertexAI});
+    this.ttsClient = new textToSpeech.TextToSpeechClient(ttsOptions);
+    this.speechClient = new speech.SpeechClient(speechOptions);
+  }
+  /**
+   * Check if Vertex AI mode is enabled
+   * @returns {boolean} True if using Vertex AI
+   */
+  isUsingVertexAI() {
+    return this.vertexAI;
+  }
+  /**
+   * Get the configured Google Cloud project ID
+   * @returns {string | undefined} The project ID or undefined if not set
+   */
+  getProject() {
+    return this.project;
+  }
+  /**
+   * Get the configured Google Cloud location/region
+   * @returns {string} The location (default: 'us-central1')
+   */
+  getLocation() {
+    return this.location;
   }
   /**
    * Gets a list of available voices
    * @returns {Promise<Array<{voiceId: string, languageCodes: string[]}>>} List of available voices and their supported languages. Default language is en-US.
    */
   async getSpeakers({ languageCode = "en-US" } = {}) {
-    return this.traced(async () => {
-      const [response] = await this.ttsClient.listVoices({ languageCode });
-      return (response?.voices || []).filter((voice) => voice.name && voice.languageCodes).map((voice) => ({
-        voiceId: voice.name,
-        languageCodes: voice.languageCodes
-      }));
-    }, "voice.google.getSpeakers")();
+    const [response] = await this.ttsClient.listVoices({ languageCode });
+    return (response?.voices || []).filter((voice) => voice.name && voice.languageCodes).map((voice) => ({
+      voiceId: voice.name,
+      languageCodes: voice.languageCodes
+    }));
   }
   async streamToString(stream) {
     const chunks = [];
@@ -83,27 +151,25 @@ var GoogleVoice = class extends voice.MastraVoice {
    * @returns {Promise<NodeJS.ReadableStream>} Stream of synthesized audio. Default encoding is LINEAR16.
    */
   async speak(input, options) {
-    return this.traced(async () => {
-      const text = typeof input === "string" ? input : await this.streamToString(input);
-      const request = {
-        input: { text },
-        voice: {
-          name: options?.speaker || this.speaker,
-          languageCode: options?.languageCode || options?.speaker?.split("-").slice(0, 2).join("-") || "en-US"
-        },
-        audioConfig: options?.audioConfig || { audioEncoding: "LINEAR16" }
-      };
-      const [response] = await this.ttsClient.synthesizeSpeech(request);
-      if (!response.audioContent) {
-        throw new Error("No audio content returned.");
-      }
-      if (typeof response.audioContent === "string") {
-        throw new Error("Audio content is a string.");
-      }
-      const stream$1 = new stream.PassThrough();
-      stream$1.end(Buffer.from(response.audioContent));
-      return stream$1;
-    }, "voice.google.speak")();
+    const text = typeof input === "string" ? input : await this.streamToString(input);
+    const request = {
+      input: { text },
+      voice: {
+        name: options?.speaker || this.speaker,
+        languageCode: options?.languageCode || options?.speaker?.split("-").slice(0, 2).join("-") || "en-US"
+      },
+      audioConfig: options?.audioConfig || { audioEncoding: "LINEAR16" }
+    };
+    const [response] = await this.ttsClient.synthesizeSpeech(request);
+    if (!response.audioContent) {
+      throw new Error("No audio content returned.");
+    }
+    if (typeof response.audioContent === "string") {
+      throw new Error("Audio content is a string.");
+    }
+    const stream$1 = new stream.PassThrough();
+    stream$1.end(Buffer.from(response.audioContent));
+    return stream$1;
   }
   /**
    * Checks if listening capabilities are enabled.
@@ -121,43 +187,39 @@ var GoogleVoice = class extends voice.MastraVoice {
    * @returns {Promise<string>} Transcribed text
    */
   async listen(audioStream, options) {
-    return this.traced(async () => {
-      const chunks = [];
-      for await (const chunk of audioStream) {
-        if (typeof chunk === "string") {
-          chunks.push(Buffer.from(chunk));
-        } else {
-          chunks.push(chunk);
-        }
+    const chunks = [];
+    for await (const chunk of audioStream) {
+      if (typeof chunk === "string") {
+        chunks.push(Buffer.from(chunk));
+      } else {
+        chunks.push(chunk);
       }
-      const buffer = Buffer.concat(chunks);
-      let request = {
-        config: {
-          encoding: "LINEAR16",
-          languageCode: "en-US",
-          ...options?.config
-        },
-        audio: {
-          content: buffer.toString("base64")
-        }
-      };
-      console.log(`BEFORE REQUEST`);
-      const [response] = await this.speechClient.recognize(request);
-      console.log(`AFTER REQUEST`);
-      if (!response.results || response.results.length === 0) {
-        throw new Error("No transcription results returned");
+    }
+    const buffer = Buffer.concat(chunks);
+    let request = {
+      config: {
+        encoding: "LINEAR16",
+        languageCode: "en-US",
+        ...options?.config
+      },
+      audio: {
+        content: buffer.toString("base64")
       }
-      const transcription = response.results.map((result) => {
-        if (!result.alternatives || result.alternatives.length === 0) {
-          return "";
-        }
-        return result.alternatives[0].transcript || "";
-      }).filter((text) => text.length > 0).join(" ");
-      if (!transcription) {
-        throw new Error("No valid transcription found in results");
+    };
+    const [response] = await this.speechClient.recognize(request);
+    if (!response.results || response.results.length === 0) {
+      throw new Error("No transcription results returned");
+    }
+    const transcription = response.results.map((result) => {
+      if (!result.alternatives || result.alternatives.length === 0) {
+        return "";
       }
-      return transcription;
-    }, "voice.google.listen")();
+      return result.alternatives[0].transcript || "";
+    }).filter((text) => text.length > 0).join(" ");
+    if (!transcription) {
+      throw new Error("No valid transcription found in results");
+    }
+    return transcription;
   }
 };
 

package/dist/index.cjs.map

@@ -1 +1 @@
-{"version":3,"sources":["../src/index.ts"],"names":["MastraVoice","TextToSpeechClient","SpeechClient","stream","PassThrough"],"mappings":";;;;;;;;AAiBA,IAAM,aAAA,GAAgB,gBAAA;AAOf,IAAM,WAAA,GAAN,cAA0BA,iBAAA,CAAY;AAAA,EACnC,SAAA;AAAA,EACA,YAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAUR,WAAA,CAAY;AAAA,IACV,cAAA;AAAA,IACA,WAAA;AAAA,IACA;AAAA,GACF,GAII,EAAC,EAAG;AACN,IAAA,MAAM,aAAA,GAAgB,QAAQ,GAAA,CAAI,cAAA;AAClC,IAAA,MAAM,cAAA,GAAiB,aAAA;AAEvB,IAAA,KAAA,CAAM;AAAA,MACJ,WAAA,EAAa;AAAA,QACX,IAAA,EAAM,EAAA;AAAA,QACN,MAAA,EAAQ,aAAa,MAAA,IAAU;AAAA,OACjC;AAAA,MACA,cAAA,EAAgB;AAAA,QACd,IAAA,EAAM,EAAA;AAAA,QACN,MAAA,EAAQ,gBAAgB,MAAA,IAAU;AAAA,OACpC;AAAA,MACA,SAAS,OAAA,IAAW;AAAA,KACrB,CAAA;AAED,IAAA,MAAM,MAAA,GAAS,aAAA,IAAiB,WAAA,EAAa,MAAA,IAAU,cAAA,EAAgB,MAAA;AACvE,IAAA,IAAI,CAAC,MAAA,EAAQ;AACX,MAAA,MAAM,IAAI,KAAA;AAAA,QACR;AAAA,OACF;AAAA,IACF;AAEA,IAAA,IAAA,CAAK,SAAA,GAAY,IAAIC,+BAAA,CAAmB;AAAA,MACtC,MAAA,EAAQ,IAAA,CAAK,WAAA,EAAa,MAAA,IAAU;AAAA,KACrC,CAAA;AAED,IAAA,IAAA,CAAK,YAAA,GAAe,IAAIC,mBAAA,CAAa;AAAA,MACnC,MAAA,EAAQ,IAAA,CAAK,cAAA,EAAgB,MAAA,IAAU;AAAA,KACxC,CAAA;AAAA,EACH;AAAA;AAAA;AAAA;AAAA;AAAA,EAMA,MAAM,WAAA,CAAY,EAAE,eAAe,OAAA,EAAQ,GAA+B,EAAC,EAAG;AAC5E,IAAA,OAAO,IAAA,CAAK,OAAO,YAAY;AAC7B,MAAA,MAAM,CAAC,QAAQ,CAAA,GAAI,MAAM,KAAK,SAAA,CAAU,UAAA,CAAW,EAAE,YAAA,EAA4B,CAAA;AACjF,MAAA,OAAA,CAAQ,QAAA,EAAU,MAAA,IAAU,EAAC,EAC1B,MAAA,CAAO,CAAA,KAAA,KAAS,KAAA,CAAM,IAAA,IAAQ,KAAA,CAAM,aAAa,CAAA,CACjD,GAAA,CAAI,CAAA,KAAA,MAAU;AAAA,QACb,SAAS,KAAA,CAAM,IAAA;AAAA,QACf,eAAe,KAAA,CAAM;AAAA,OACvB,CAAE,CAAA;AAAA,IACN,CAAA,EAAG,0BAA0B,CAAA,EAAE;AAAA,EACjC;AAAA,EAEA,MAAc,eAAe,MAAA,EAAgD;AAC3E,IAAA,MAAM,SAAmB,EAAC;AAC1B,IAAA,WAAA,MAAiB,SAAS,MAAA,EAAQ;AAChC,MAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,QAAA,MAAA,CAAO,IAAA,CAAK,MAAA,CAAO,IAAA,CAAK,KAAK,CAAC,CAAA;AAAA,MAChC,CAAA,MAAO;AACL,QAAA,MAAA,CAAO,KAAK,KAAK,CAAA;AAAA,MACnB;AAAA,IACF;AACA,IAAA,OAAO,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA,CAAE,SAAS,OAAO,CAAA;AAAA,EAC/C;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAWA,MAAM,KAAA,CACJ,KAAA,EACA,OAAA,EAKgC;AAChC,IAAA,OAAO,IAAA,CAAK,OAAO,YAAY;AAC7B,MAAA,MAAM,IAAA,GAAO,OAAO,KAAA,KAAU,QAAA,GAAW,QAAQ,MAAM,IAAA,CAAK,eAAe,KAAK,CAAA;AAEhF,MAAA,MAAM,OAAA,GAA4E;AAAA,QAChF,KAAA,EAAO,EAAE,IAAA,EAAK;AAAA,QACd,KAAA,EAAO;AAAA,UACL,IAAA,EAAM,OAAA,EAAS,OAAA,IAAW,IAAA,CAAK,OAAA;AAAA,UAC/B,YAAA,EAAc,OAAA,EAAS,YAAA,IAAgB,OAAA,EAAS,SAAS,KAAA,CAAM,GAAG,CAAA,CAAE,KAAA,CAAM,CAAA,EAAG,CAAC,CAAA,CAAE,IAAA,CAAK,GAAG,CAAA,IAAK;AAAA,SAC/F;AAAA,QACA,WAAA,EAAa,OAAA,EAAS,WAAA,IAAe,EAAE,eAAe,UAAA;AAAW,OACnE;AAEA,MAAA,MAAM,CAAC,QAAQ,CAAA,GAAI,MAAM,IAAA,CAAK,SAAA,CAAU,iBAAiB,OAAO,CAAA;AAEhE,MAAA,IAAI,CAAC,SAAS,YAAA,EAAc;AAC1B,QAAA,MAAM,IAAI,MAAM,4BAA4B,CAAA;AAAA,MAC9C;AAEA,MAAA,IAAI,OAAO,QAAA,CAAS,YAAA,KAAiB,QAAA,EAAU;AAC7C,QAAA,MAAM,IAAI,MAAM,4BAA4B,CAAA;AAAA,MAC9C;AAEA,MAAA,MAAMC,QAAA,GAAS,IAAIC,kBAAA,EAAY;AAC/B,MAAAD,QAAA,CAAO,GAAA,CAAI,MAAA,CAAO,IAAA,CAAK,QAAA,CAAS,YAAY,CAAC,CAAA;AAC7C,MAAA,OAAOA,QAAA;AAAA,IACT,CAAA,EAAG,oBAAoB,CAAA,EAAE;AAAA,EAC3B;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAOA,MAAM,WAAA,GAAc;AAClB,IAAA,OAAO,EAAE,SAAS,IAAA,EAAK;AAAA,EACzB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EASA,MAAM,MAAA,CACJ,WAAA,EACA,OAAA,EACiB;AACjB,IAAA,OAAO,IAAA,CAAK,OAAO,YAAY;AAC7B,MAAA,MAAM,SAAmB,EAAC;AAC1B,MAAA,WAAA,MAAiB,SAAS,WAAA,EAAa;AACrC,QAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,UAAA,MAAA,CAAO,IAAA,CAAK,MAAA,CAAO,IAAA,CAAK,KAAK,CAAC,CAAA;AAAA,QAChC,CAAA,MAAO;AACL,UAAA,MAAA,CAAO,KAAK,KAAK,CAAA;AAAA,QACnB;AAAA,MACF;AACA,MAAA,MAAM,MAAA,GAAS,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA;AAEnC,MAAA,IAAI,OAAA,GAAU;AAAA,QACZ,MAAA,EAAQ;AAAA,UACN,QAAA,EAAU,UAAA;AAAA,UACV,YAAA,EAAc,OAAA;AAAA,UACd,GAAG,OAAA,EAAS;AAAA,SACd;AAAA,QACA,KAAA,EAAO;AAAA,UACL,OAAA,EAAS,MAAA,CAAO,QAAA,CAAS,QAAQ;AAAA;AACnC,OACF;AACA,MAAA,OAAA,CAAQ,IAAI,CAAA,cAAA,CAAgB,CAAA;AAC5B,MAAA,MAAM,CAAC,QAAQ,CAAA,GAAI,MAAM,IAAA,CAAK,YAAA,CAAa,UAAU,OAAwD,CAAA;AAC7G,MAAA,OAAA,CAAQ,IAAI,CAAA,aAAA,CAAe,CAAA;AAE3B,MAAA,IAAI,CAAC,QAAA,CAAS,OAAA,IAAW,QAAA,CAAS,OAAA,CAAQ,WAAW,CAAA,EAAG;AACtD,QAAA,MAAM,IAAI,MAAM,mCAAmC,CAAA;AAAA,MACrD;AAEA,MAAA,MAAM,aAAA,GAAgB,QAAA,CAAS,OAAA,CAC5B,GAAA,CAAI,CAAC,MAAA,KAAgB;AACpB,QAAA,IAAI,CAAC,MAAA,CAAO,YAAA,IAAgB,MAAA,CAAO,YAAA,CAAa,WAAW,CAAA,EAAG;AAC5D,UAAA,OAAO,EAAA;AAAA,QACT;AACA,QAAA,OAAO,MAAA,CAAO,YAAA,CAAa,CAAC,CAAA,CAAE,UAAA,IAAc,EAAA;AAAA,MAC9C,CAAC,CAAA,CACA,MAAA,CAAO,CAAC,IAAA,KAAiB,KAAK,MAAA,GAAS,CAAC,CAAA,CACxC,IAAA,CAAK,GAAG,CAAA;AAEX,MAAA,IAAI,CAAC,aAAA,EAAe;AAClB,QAAA,MAAM,IAAI,MAAM,yCAAyC,CAAA;AAAA,MAC3D;AAEA,MAAA,OAAO,aAAA;AAAA,IACT,CAAA,EAAG,qBAAqB,CAAA,EAAE;AAAA,EAC5B;AACF","file":"index.cjs","sourcesContent":["import { PassThrough } from 'stream';\n\nimport { SpeechClient } from '@google-cloud/speech';\nimport type { google as SpeechTypes } from '@google-cloud/speech/build/protos/protos';\nimport { TextToSpeechClient } from '@google-cloud/text-to-speech';\nimport type { google as TextToSpeechTypes } from '@google-cloud/text-to-speech/build/protos/protos';\nimport { MastraVoice } from '@mastra/core/voice';\n\n/**\n * Configuration for Google Cloud Voice models\n * @interface GoogleModelConfig\n * @property {string} [apiKey] - Optional Google Cloud API key. If not provided, will use GOOGLE_API_KEY environment variable\n */\nexport interface GoogleModelConfig {\n  apiKey?: string;\n}\n\nconst DEFAULT_VOICE = 'en-US-Casual-K';\n\n/**\n * GoogleVoice class provides Text-to-Speech and Speech-to-Text capabilities using Google Cloud services\n * @class GoogleVoice\n * @extends MastraVoice\n */\nexport class GoogleVoice extends MastraVoice {\n  private ttsClient: TextToSpeechClient;\n  private speechClient: SpeechClient;\n\n  /**\n   * Creates an instance of GoogleVoice\n   * @param {Object} config - Configuration options\n   * @param {GoogleModelConfig} [config.speechModel] - Configuration for speech synthesis\n   * @param {GoogleModelConfig} [config.listeningModel] - Configuration for speech recognition\n   * @param {string} [config.speaker] - Default voice ID to use for speech synthesis\n   * @throws {Error} If no API key is provided via config or environment variable\n   */\n  constructor({\n    listeningModel,\n    speechModel,\n    speaker,\n  }: {\n    listeningModel?: GoogleModelConfig;\n    speechModel?: GoogleModelConfig;\n    speaker?: string;\n  } = {}) {\n    const defaultApiKey = process.env.GOOGLE_API_KEY;\n    const defaultSpeaker = DEFAULT_VOICE;\n\n    super({\n      speechModel: {\n        name: '',\n        apiKey: speechModel?.apiKey ?? defaultApiKey,\n      },\n      listeningModel: {\n        name: '',\n        apiKey: listeningModel?.apiKey ?? defaultApiKey,\n      },\n      speaker: speaker ?? defaultSpeaker,\n    });\n\n    const apiKey = defaultApiKey || speechModel?.apiKey || listeningModel?.apiKey;\n    if (!apiKey) {\n      throw new Error(\n        'Google API key is not set, set GOOGLE_API_KEY environment variable or pass apiKey to constructor',\n      );\n    }\n\n    this.ttsClient = new TextToSpeechClient({\n      apiKey: this.speechModel?.apiKey || defaultApiKey,\n    });\n\n    this.speechClient = new SpeechClient({\n      apiKey: this.listeningModel?.apiKey || defaultApiKey,\n    });\n  }\n\n  /**\n   * Gets a list of available voices\n   * @returns {Promise<Array<{voiceId: string, languageCodes: string[]}>>} List of available voices and their supported languages. Default language is en-US.\n   */\n  async getSpeakers({ languageCode = 'en-US' }: { languageCode?: string } = {}) {\n    return this.traced(async () => {\n      const [response] = await this.ttsClient.listVoices({ languageCode: languageCode });\n      return (response?.voices || [])\n        .filter(voice => voice.name && voice.languageCodes)\n        .map(voice => ({\n          voiceId: voice.name!,\n          languageCodes: voice.languageCodes!,\n        }));\n    }, 'voice.google.getSpeakers')();\n  }\n\n  private async streamToString(stream: NodeJS.ReadableStream): Promise<string> {\n    const chunks: Buffer[] = [];\n    for await (const chunk of stream) {\n      if (typeof chunk === 'string') {\n        chunks.push(Buffer.from(chunk));\n      } else {\n        chunks.push(chunk);\n      }\n    }\n    return Buffer.concat(chunks).toString('utf-8');\n  }\n\n  /**\n   * Converts text to speech\n   * @param {string | NodeJS.ReadableStream} input - Text or stream to convert to speech\n   * @param {Object} [options] - Speech synthesis options\n   * @param {string} [options.speaker] - Voice ID to use\n   * @param {string} [options.languageCode] - Language code for the voice\n   * @param {TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig']} [options.audioConfig] - Audio configuration options\n   * @returns {Promise<NodeJS.ReadableStream>} Stream of synthesized audio. Default encoding is LINEAR16.\n   */\n  async speak(\n    input: string | NodeJS.ReadableStream,\n    options?: {\n      speaker?: string;\n      languageCode?: string;\n      audioConfig?: TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig'];\n    },\n  ): Promise<NodeJS.ReadableStream> {\n    return this.traced(async () => {\n      const text = typeof input === 'string' ? input : await this.streamToString(input);\n\n      const request: TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest = {\n        input: { text },\n        voice: {\n          name: options?.speaker || this.speaker,\n          languageCode: options?.languageCode || options?.speaker?.split('-').slice(0, 2).join('-') || 'en-US',\n        },\n        audioConfig: options?.audioConfig || { audioEncoding: 'LINEAR16' },\n      };\n\n      const [response] = await this.ttsClient.synthesizeSpeech(request);\n\n      if (!response.audioContent) {\n        throw new Error('No audio content returned.');\n      }\n\n      if (typeof response.audioContent === 'string') {\n        throw new Error('Audio content is a string.');\n      }\n\n      const stream = new PassThrough();\n      stream.end(Buffer.from(response.audioContent));\n      return stream;\n    }, 'voice.google.speak')();\n  }\n\n  /**\n   * Checks if listening capabilities are enabled.\n   *\n   * @returns {Promise<{ enabled: boolean }>}\n   */\n  async getListener() {\n    return { enabled: true };\n  }\n\n  /**\n   * Converts speech to text\n   * @param {NodeJS.ReadableStream} audioStream - Audio stream to transcribe. Default encoding is LINEAR16.\n   * @param {Object} [options] - Recognition options\n   * @param {SpeechTypes.cloud.speech.v1.IRecognitionConfig} [options.config] - Recognition configuration\n   * @returns {Promise<string>} Transcribed text\n   */\n  async listen(\n    audioStream: NodeJS.ReadableStream,\n    options?: { stream?: boolean; config?: SpeechTypes.cloud.speech.v1.IRecognitionConfig },\n  ): Promise<string> {\n    return this.traced(async () => {\n      const chunks: Buffer[] = [];\n      for await (const chunk of audioStream) {\n        if (typeof chunk === 'string') {\n          chunks.push(Buffer.from(chunk));\n        } else {\n          chunks.push(chunk);\n        }\n      }\n      const buffer = Buffer.concat(chunks);\n\n      let request = {\n        config: {\n          encoding: 'LINEAR16',\n          languageCode: 'en-US',\n          ...options?.config,\n        },\n        audio: {\n          content: buffer.toString('base64'),\n        },\n      };\n      console.log(`BEFORE REQUEST`);\n      const [response] = await this.speechClient.recognize(request as SpeechTypes.cloud.speech.v1.IRecognizeRequest);\n      console.log(`AFTER REQUEST`);\n\n      if (!response.results || response.results.length === 0) {\n        throw new Error('No transcription results returned');\n      }\n\n      const transcription = response.results\n        .map((result: any) => {\n          if (!result.alternatives || result.alternatives.length === 0) {\n            return '';\n          }\n          return result.alternatives[0].transcript || '';\n        })\n        .filter((text: string) => text.length > 0)\n        .join(' ');\n\n      if (!transcription) {\n        throw new Error('No valid transcription found in results');\n      }\n\n      return transcription;\n    }, 'voice.google.listen')();\n  }\n}\n"]}
+{"version":3,"sources":["../src/index.ts"],"names":["MastraVoice","TextToSpeechClient","SpeechClient","stream","PassThrough"],"mappings":";;;;;;;;AA6DA,IAAM,iBAAA,GAAoB,CACxB,WAAA,EACA,QAAA,EACA,YAAA,KACe;AACf,EAAA,MAAM,WAAuB,EAAC;AAG9B,EAAA,IAAI,cAAc,QAAA,EAAU;AAC1B,IAAA,MAAM,SAAA,GAAY,YAAA,CAAa,OAAA,IAAW,OAAA,CAAQ,GAAA,CAAI,oBAAA;AACtD,IAAA,IAAI,SAAA,EAAW;AACb,MAAA,QAAA,CAAS,SAAA,GAAY,SAAA;AAAA,IACvB;AAAA,EACF;AAEA,EAAA,MAAM,MAAA,GAAS,WAAA,EAAa,MAAA,IAAU,QAAA,CAAS,MAAA;AAE/C,EAAA,IAAI,MAAA,IAAU,CAAC,YAAA,EAAc,QAAA,EAAU;AACrC,IAAA,QAAA,CAAS,MAAA,GAAS,MAAA;AAAA,EACpB;AAEA,EAAA,MAAM,WAAA,GAAc,WAAA,EAAa,WAAA,IAAe,QAAA,CAAS,WAAA;AACzD,EAAA,IAAI,WAAA,EAAa;AACf,IAAA,QAAA,CAAS,WAAA,GAAc,WAAA;AAAA,EACzB;AAEA,EAAA,MAAM,WAAA,GAAc,WAAA,EAAa,WAAA,IAAe,QAAA,CAAS,WAAA;AACzD,EAAA,IAAI,WAAA,EAAa;AACf,IAAA,QAAA,CAAS,WAAA,GAAc,WAAA;AAAA,EACzB;AAEA,EAAA,OAAO,QAAA;AACT,CAAA;AAEA,IAAM,gBAAA,GAAmB,CACvB,MAAA,EACA,YAAA,KACwB;AACxB,EAAA,MAAM,UAA+B,EAAC;AAEtC,EAAA,IAAI,OAAO,WAAA,EAAa;AACtB,IAAA,OAAA,CAAQ,cAAc,MAAA,CAAO,WAAA;AAAA,EAC/B;AAEA,EAAA,IAAI,OAAO,WAAA,EAAa;AACtB,IAAA,OAAA,CAAQ,cAAc,MAAA,CAAO,WAAA;AAAA,EAC/B;AAGA,EAAA,IAAI,MAAA,CAAO,MAAA,IAAU,CAAC,YAAA,EAAc,QAAA,EAAU;AAC5C,IAAA,OAAA,CAAQ,SAAS,MAAA,CAAO,MAAA;AAAA,EAC1B;AAGA,EAAA,IAAI,OAAO,SAAA,EAAW;AACpB,IAAA,OAAA,CAAQ,YAAY,MAAA,CAAO,SAAA;AAAA,EAC7B;AAEA,EAAA,OAAO,OAAA;AACT,CAAA;AAEA,IAAM,aAAA,GAAgB,gBAAA;AAsCf,IAAM,WAAA,GAAN,cAA0BA,iBAAA,CAAY;AAAA,EACnC,SAAA;AAAA,EACA,YAAA;AAAA,EACS,QAAA;AAAA,EACA,OAAA;AAAA,EACA,QAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYjB,WAAA,CAAY,EAAE,cAAA,EAAgB,WAAA,EAAa,OAAA,EAAS,QAAA,GAAW,KAAA,EAAO,OAAA,EAAS,QAAA,EAAS,GAAuB,EAAC,EAAG;AACjH,IAAA,MAAM,aAAA,GAAgB,QAAQ,GAAA,CAAI,cAAA;AAClC,IAAA,MAAM,kBAAA,GAAqB,QAAQ,GAAA,CAAI,8BAAA;AACvC,IAAA,MAAM,cAAA,GAAiB,aAAA;AAGvB,IAAA,MAAM,eAAA,GAAkB,OAAA,IAAW,OAAA,CAAQ,GAAA,CAAI,oBAAA;AAC/C,IAAA,MAAM,gBAAA,GAAmB,QAAA,IAAY,OAAA,CAAQ,GAAA,CAAI,qBAAA,IAAyB,aAAA;AAG1E,IAAA,IAAI,QAAA,IAAY,CAAC,eAAA,EAAiB;AAChC,MAAA,MAAM,IAAI,KAAA;AAAA,QACR;AAAA,OAEF;AAAA,IACF;AAEA,IAAA,MAAM,YAAA,GAAe,EAAE,QAAA,EAAU,OAAA,EAAS,eAAA,EAAgB;AAE1D,IAAA,MAAM,cAAA,GAA6B;AAAA,MACjC,MAAA,EAAQ,aAAA,IAAiB,WAAA,EAAa,MAAA,IAAU,cAAA,EAAgB,MAAA;AAAA,MAChE,WAAA,EAAa,kBAAA,IAAsB,WAAA,EAAa,WAAA,IAAe,cAAA,EAAgB,WAAA;AAAA,MAC/E,WAAA,EAAa,WAAA,EAAa,WAAA,IAAe,cAAA,EAAgB,WAE3D,CAAA;AAEA,IAAA,MAAM,gBAAA,GAAmB,iBAAA,CAAkB,WAAA,EAAa,cAAA,EAAgB,YAAY,CAAA;AACpF,IAAA,MAAM,mBAAA,GAAsB,iBAAA,CAAkB,cAAA,EAAgB,cAAA,EAAgB,YAAY,CAAA;AAE1F,IAAA,KAAA,CAAM;AAAA,MACJ,WAAA,EAAa;AAAA,QACX,IAAA,EAAM,EAAA;AAAA,QACN,MAAA,EAAQ,iBAAiB,MAAA,IAAU;AAAA,OACrC;AAAA,MACA,cAAA,EAAgB;AAAA,QACd,IAAA,EAAM,EAAA;AAAA,QACN,MAAA,EAAQ,oBAAoB,MAAA,IAAU;AAAA,OACxC;AAAA,MACA,SAAS,OAAA,IAAW;AAAA,KACrB,CAAA;AAED,IAAA,IAAA,CAAK,QAAA,GAAW,QAAA;AAChB,IAAA,IAAA,CAAK,OAAA,GAAU,eAAA;AACf,IAAA,IAAA,CAAK,QAAA,GAAW,gBAAA;AAEhB,IAAA,MAAM,aAAa,gBAAA,CAAiB,gBAAA,EAAkB,EAAE,QAAqC,CAAC,CAAA;AAC9F,IAAA,MAAM,gBAAgB,gBAAA,CAAiB,mBAAA,EAAqB,EAAE,QAAqC,CAAC,CAAA;AAEpG,IAAA,IAAA,CAAK,SAAA,GAAY,IAAIC,+BAAA,CAAmB,UAAU,CAAA;AAClD,IAAA,IAAA,CAAK,YAAA,GAAe,IAAIC,mBAAA,CAAa,aAAa,CAAA;AAAA,EACpD;AAAA;AAAA;AAAA;AAAA;AAAA,EAMA,eAAA,GAA2B;AACzB,IAAA,OAAO,IAAA,CAAK,QAAA;AAAA,EACd;AAAA;AAAA;AAAA;AAAA;AAAA,EAMA,UAAA,GAAiC;AAC/B,IAAA,OAAO,IAAA,CAAK,OAAA;AAAA,EACd;AAAA;AAAA;AAAA;AAAA;AAAA,EAMA,WAAA,GAAsB;AACpB,IAAA,OAAO,IAAA,CAAK,QAAA;AAAA,EACd;AAAA;AAAA;AAAA;AAAA;AAAA,EAMA,MAAM,WAAA,CAAY,EAAE,eAAe,OAAA,EAAQ,GAA+B,EAAC,EAAG;AAC5E,IAAA,MAAM,CAAC,QAAQ,CAAA,GAAI,MAAM,KAAK,SAAA,CAAU,UAAA,CAAW,EAAE,YAAA,EAA4B,CAAA;AACjF,IAAA,OAAA,CAAQ,QAAA,EAAU,MAAA,IAAU,EAAC,EAC1B,MAAA,CAAO,CAAA,KAAA,KAAS,KAAA,CAAM,IAAA,IAAQ,KAAA,CAAM,aAAa,CAAA,CACjD,GAAA,CAAI,CAAA,KAAA,MAAU;AAAA,MACb,SAAS,KAAA,CAAM,IAAA;AAAA,MACf,eAAe,KAAA,CAAM;AAAA,KACvB,CAAE,CAAA;AAAA,EACN;AAAA,EAEA,MAAc,eAAe,MAAA,EAAgD;AAC3E,IAAA,MAAM,SAAmB,EAAC;AAC1B,IAAA,WAAA,MAAiB,SAAS,MAAA,EAAQ;AAChC,MAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,QAAA,MAAA,CAAO,IAAA,CAAK,MAAA,CAAO,IAAA,CAAK,KAAK,CAAC,CAAA;AAAA,MAChC,CAAA,MAAO;AACL,QAAA,MAAA,CAAO,KAAK,KAAK,CAAA;AAAA,MACnB;AAAA,IACF;AACA,IAAA,OAAO,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA,CAAE,SAAS,OAAO,CAAA;AAAA,EAC/C;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAWA,MAAM,KAAA,CACJ,KAAA,EACA,OAAA,EAKgC;AAChC,IAAA,MAAM,IAAA,GAAO,OAAO,KAAA,KAAU,QAAA,GAAW,QAAQ,MAAM,IAAA,CAAK,eAAe,KAAK,CAAA;AAEhF,IAAA,MAAM,OAAA,GAA4E;AAAA,MAChF,KAAA,EAAO,EAAE,IAAA,EAAK;AAAA,MACd,KAAA,EAAO;AAAA,QACL,IAAA,EAAM,OAAA,EAAS,OAAA,IAAW,IAAA,CAAK,OAAA;AAAA,QAC/B,YAAA,EAAc,OAAA,EAAS,YAAA,IAAgB,OAAA,EAAS,SAAS,KAAA,CAAM,GAAG,CAAA,CAAE,KAAA,CAAM,CAAA,EAAG,CAAC,CAAA,CAAE,IAAA,CAAK,GAAG,CAAA,IAAK;AAAA,OAC/F;AAAA,MACA,WAAA,EAAa,OAAA,EAAS,WAAA,IAAe,EAAE,eAAe,UAAA;AAAW,KACnE;AAEA,IAAA,MAAM,CAAC,QAAQ,CAAA,GAAI,MAAM,IAAA,CAAK,SAAA,CAAU,iBAAiB,OAAO,CAAA;AAEhE,IAAA,IAAI,CAAC,SAAS,YAAA,EAAc;AAC1B,MAAA,MAAM,IAAI,MAAM,4BAA4B,CAAA;AAAA,IAC9C;AAEA,IAAA,IAAI,OAAO,QAAA,CAAS,YAAA,KAAiB,QAAA,EAAU;AAC7C,MAAA,MAAM,IAAI,MAAM,4BAA4B,CAAA;AAAA,IAC9C;AAEA,IAAA,MAAMC,QAAA,GAAS,IAAIC,kBAAA,EAAY;AAC/B,IAAAD,QAAA,CAAO,GAAA,CAAI,MAAA,CAAO,IAAA,CAAK,QAAA,CAAS,YAAY,CAAC,CAAA;AAC7C,IAAA,OAAOA,QAAA;AAAA,EACT;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAOA,MAAM,WAAA,GAAc;AAClB,IAAA,OAAO,EAAE,SAAS,IAAA,EAAK;AAAA,EACzB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EASA,MAAM,MAAA,CACJ,WAAA,EACA,OAAA,EACiB;AACjB,IAAA,MAAM,SAAmB,EAAC;AAC1B,IAAA,WAAA,MAAiB,SAAS,WAAA,EAAa;AACrC,MAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,QAAA,MAAA,CAAO,IAAA,CAAK,MAAA,CAAO,IAAA,CAAK,KAAK,CAAC,CAAA;AAAA,MAChC,CAAA,MAAO;AACL,QAAA,MAAA,CAAO,KAAK,KAAK,CAAA;AAAA,MACnB;AAAA,IACF;AACA,IAAA,MAAM,MAAA,GAAS,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA;AAEnC,IAAA,IAAI,OAAA,GAAU;AAAA,MACZ,MAAA,EAAQ;AAAA,QACN,QAAA,EAAU,UAAA;AAAA,QACV,YAAA,EAAc,OAAA;AAAA,QACd,GAAG,OAAA,EAAS;AAAA,OACd;AAAA,MACA,KAAA,EAAO;AAAA,QACL,OAAA,EAAS,MAAA,CAAO,QAAA,CAAS,QAAQ;AAAA;AACnC,KACF;AACA,IAAA,MAAM,CAAC,QAAQ,CAAA,GAAI,MAAM,IAAA,CAAK,YAAA,CAAa,UAAU,OAAwD,CAAA;AAE7G,IAAA,IAAI,CAAC,QAAA,CAAS,OAAA,IAAW,QAAA,CAAS,OAAA,CAAQ,WAAW,CAAA,EAAG;AACtD,MAAA,MAAM,IAAI,MAAM,mCAAmC,CAAA;AAAA,IACrD;AAEA,IAAA,MAAM,aAAA,GAAgB,QAAA,CAAS,OAAA,CAC5B,GAAA,CAAI,CAAC,MAAA,KAAgB;AACpB,MAAA,IAAI,CAAC,MAAA,CAAO,YAAA,IAAgB,MAAA,CAAO,YAAA,CAAa,WAAW,CAAA,EAAG;AAC5D,QAAA,OAAO,EAAA;AAAA,MACT;AACA,MAAA,OAAO,MAAA,CAAO,YAAA,CAAa,CAAC,CAAA,CAAE,UAAA,IAAc,EAAA;AAAA,IAC9C,CAAC,CAAA,CACA,MAAA,CAAO,CAAC,IAAA,KAAiB,KAAK,MAAA,GAAS,CAAC,CAAA,CACxC,IAAA,CAAK,GAAG,CAAA;AAEX,IAAA,IAAI,CAAC,aAAA,EAAe;AAClB,MAAA,MAAM,IAAI,MAAM,yCAAyC,CAAA;AAAA,IAC3D;AAEA,IAAA,OAAO,aAAA;AAAA,EACT;AACF","file":"index.cjs","sourcesContent":["import { PassThrough } from 'node:stream';\n\nimport { SpeechClient } from '@google-cloud/speech';\nimport type { google as SpeechTypes } from '@google-cloud/speech/build/protos/protos';\nimport { TextToSpeechClient } from '@google-cloud/text-to-speech';\nimport type { google as TextToSpeechTypes } from '@google-cloud/text-to-speech/build/protos/protos';\nimport { MastraVoice } from '@mastra/core/voice';\n\n/**\n * Configuration for Google Cloud Voice models\n * @interface GoogleModelConfig\n * @property {string} [apiKey] - Optional Google Cloud API key. If not provided, will use GOOGLE_API_KEY environment variable\n * @property {string} [keyFilename] - Optional path to a service account key file. If not provided, will use GOOGLE_APPLICATION_CREDENTIALS environment variable\n * @property {{ client_email?: string; private_key?: string }} [credentials] - Optional in-memory service account credentials\n */\nexport interface GoogleModelConfig {\n  apiKey?: string;\n  keyFilename?: string;\n  credentials?: {\n    client_email?: string;\n    private_key?: string;\n    [key: string]: unknown;\n  };\n}\n\n/**\n * Configuration options for GoogleVoice\n * @interface GoogleVoiceConfig\n */\nexport interface GoogleVoiceConfig {\n  /** Configuration for speech synthesis (TTS) */\n  speechModel?: GoogleModelConfig;\n  /** Configuration for speech recognition (STT) */\n  listeningModel?: GoogleModelConfig;\n  /** Default voice ID to use for speech synthesis */\n  speaker?: string;\n  /**\n   * Enable Vertex AI mode for enterprise deployments.\n   * When true, uses Google Cloud project-based authentication instead of API keys.\n   * Requires `project` to be set or GOOGLE_CLOUD_PROJECT environment variable.\n   */\n  vertexAI?: boolean;\n  /**\n   * Google Cloud project ID (required when vertexAI is true).\n   * Falls back to GOOGLE_CLOUD_PROJECT environment variable.\n   */\n  project?: string;\n  /**\n   * Google Cloud region for Vertex AI endpoints.\n   * Falls back to GOOGLE_CLOUD_LOCATION environment variable.\n   * @default 'us-central1'\n   */\n  location?: string;\n}\n\ntype AuthConfig = Pick<GoogleModelConfig, 'apiKey' | 'keyFilename' | 'credentials'> & {\n  projectId?: string;\n};\n\ntype GoogleClientOptions = AuthConfig;\n\nconst resolveAuthConfig = (\n  modelConfig: GoogleModelConfig | undefined,\n  fallback: AuthConfig,\n  vertexConfig?: { vertexAI?: boolean; project?: string },\n): AuthConfig => {\n  const resolved: AuthConfig = {};\n\n  // For Vertex AI, prioritize project-based auth over API keys\n  if (vertexConfig?.vertexAI) {\n    const projectId = vertexConfig.project || process.env.GOOGLE_CLOUD_PROJECT;\n    if (projectId) {\n      resolved.projectId = projectId;\n    }\n  }\n\n  const apiKey = modelConfig?.apiKey ?? fallback.apiKey;\n  // Only use API key if not in Vertex AI mode\n  if (apiKey && !vertexConfig?.vertexAI) {\n    resolved.apiKey = apiKey;\n  }\n\n  const keyFilename = modelConfig?.keyFilename ?? fallback.keyFilename;\n  if (keyFilename) {\n    resolved.keyFilename = keyFilename;\n  }\n\n  const credentials = modelConfig?.credentials ?? fallback.credentials;\n  if (credentials) {\n    resolved.credentials = credentials;\n  }\n\n  return resolved;\n};\n\nconst buildAuthOptions = (\n  config: AuthConfig,\n  vertexConfig?: { vertexAI?: boolean; location?: string },\n): GoogleClientOptions => {\n  const options: GoogleClientOptions = {};\n\n  if (config.credentials) {\n    options.credentials = config.credentials;\n  }\n\n  if (config.keyFilename) {\n    options.keyFilename = config.keyFilename;\n  }\n\n  // Only use API key if not using Vertex AI\n  if (config.apiKey && !vertexConfig?.vertexAI) {\n    options.apiKey = config.apiKey;\n  }\n\n  // For Vertex AI, set the project ID\n  if (config.projectId) {\n    options.projectId = config.projectId;\n  }\n\n  return options;\n};\n\nconst DEFAULT_VOICE = 'en-US-Casual-K';\n\n/**\n * GoogleVoice class provides Text-to-Speech and Speech-to-Text capabilities using Google Cloud services.\n * Supports both standard Google Cloud API authentication and Vertex AI mode for enterprise deployments.\n *\n * @class GoogleVoice\n * @extends MastraVoice\n *\n * @example Standard usage with API key\n * ```typescript\n * const voice = new GoogleVoice({\n *   speechModel: { apiKey: 'your-api-key' },\n *   speaker: 'en-US-Studio-O',\n * });\n * ```\n *\n * @example Vertex AI mode (recommended for production)\n * ```typescript\n * const voice = new GoogleVoice({\n *   vertexAI: true,\n *   project: 'your-gcp-project',\n *   location: 'us-central1',\n *   speaker: 'en-US-Studio-O',\n * });\n * ```\n *\n * @example Vertex AI with service account\n * ```typescript\n * const voice = new GoogleVoice({\n *   vertexAI: true,\n *   project: 'your-gcp-project',\n *   speechModel: {\n *     keyFilename: '/path/to/service-account.json',\n *   },\n * });\n * ```\n */\nexport class GoogleVoice extends MastraVoice {\n  private ttsClient: TextToSpeechClient;\n  private speechClient: SpeechClient;\n  private readonly vertexAI: boolean;\n  private readonly project?: string;\n  private readonly location: string;\n\n  /**\n   * Creates an instance of GoogleVoice\n   * @param {GoogleVoiceConfig} config - Configuration options\n   * @param {GoogleModelConfig} [config.speechModel] - Configuration for speech synthesis\n   * @param {GoogleModelConfig} [config.listeningModel] - Configuration for speech recognition\n   * @param {string} [config.speaker] - Default voice ID to use for speech synthesis\n   * @param {boolean} [config.vertexAI] - Enable Vertex AI mode\n   * @param {string} [config.project] - Google Cloud project ID (required for Vertex AI)\n   * @param {string} [config.location] - Google Cloud region (default: 'us-central1')\n   */\n  constructor({ listeningModel, speechModel, speaker, vertexAI = false, project, location }: GoogleVoiceConfig = {}) {\n    const defaultApiKey = process.env.GOOGLE_API_KEY;\n    const defaultKeyFilename = process.env.GOOGLE_APPLICATION_CREDENTIALS;\n    const defaultSpeaker = DEFAULT_VOICE;\n\n    // Resolve Vertex AI configuration\n    const resolvedProject = project || process.env.GOOGLE_CLOUD_PROJECT;\n    const resolvedLocation = location || process.env.GOOGLE_CLOUD_LOCATION || 'us-central1';\n\n    // Validate Vertex AI configuration\n    if (vertexAI && !resolvedProject) {\n      throw new Error(\n        'Google Cloud project ID is required when using Vertex AI. ' +\n          'Set GOOGLE_CLOUD_PROJECT environment variable or pass project to constructor.',\n      );\n    }\n\n    const vertexConfig = { vertexAI, project: resolvedProject };\n\n    const sharedFallback: AuthConfig = {\n      apiKey: defaultApiKey ?? speechModel?.apiKey ?? listeningModel?.apiKey,\n      keyFilename: defaultKeyFilename ?? speechModel?.keyFilename ?? listeningModel?.keyFilename,\n      credentials: speechModel?.credentials ?? listeningModel?.credentials,\n      projectId: resolvedProject,\n    };\n\n    const speechAuthConfig = resolveAuthConfig(speechModel, sharedFallback, vertexConfig);\n    const listeningAuthConfig = resolveAuthConfig(listeningModel, sharedFallback, vertexConfig);\n\n    super({\n      speechModel: {\n        name: '',\n        apiKey: speechAuthConfig.apiKey ?? defaultApiKey,\n      },\n      listeningModel: {\n        name: '',\n        apiKey: listeningAuthConfig.apiKey ?? defaultApiKey,\n      },\n      speaker: speaker ?? defaultSpeaker,\n    });\n\n    this.vertexAI = vertexAI;\n    this.project = resolvedProject;\n    this.location = resolvedLocation;\n\n    const ttsOptions = buildAuthOptions(speechAuthConfig, { vertexAI, location: resolvedLocation });\n    const speechOptions = buildAuthOptions(listeningAuthConfig, { vertexAI, location: resolvedLocation });\n\n    this.ttsClient = new TextToSpeechClient(ttsOptions);\n    this.speechClient = new SpeechClient(speechOptions);\n  }\n\n  /**\n   * Check if Vertex AI mode is enabled\n   * @returns {boolean} True if using Vertex AI\n   */\n  isUsingVertexAI(): boolean {\n    return this.vertexAI;\n  }\n\n  /**\n   * Get the configured Google Cloud project ID\n   * @returns {string | undefined} The project ID or undefined if not set\n   */\n  getProject(): string | undefined {\n    return this.project;\n  }\n\n  /**\n   * Get the configured Google Cloud location/region\n   * @returns {string} The location (default: 'us-central1')\n   */\n  getLocation(): string {\n    return this.location;\n  }\n\n  /**\n   * Gets a list of available voices\n   * @returns {Promise<Array<{voiceId: string, languageCodes: string[]}>>} List of available voices and their supported languages. Default language is en-US.\n   */\n  async getSpeakers({ languageCode = 'en-US' }: { languageCode?: string } = {}) {\n    const [response] = await this.ttsClient.listVoices({ languageCode: languageCode });\n    return (response?.voices || [])\n      .filter(voice => voice.name && voice.languageCodes)\n      .map(voice => ({\n        voiceId: voice.name!,\n        languageCodes: voice.languageCodes!,\n      }));\n  }\n\n  private async streamToString(stream: NodeJS.ReadableStream): Promise<string> {\n    const chunks: Buffer[] = [];\n    for await (const chunk of stream) {\n      if (typeof chunk === 'string') {\n        chunks.push(Buffer.from(chunk));\n      } else {\n        chunks.push(chunk);\n      }\n    }\n    return Buffer.concat(chunks).toString('utf-8');\n  }\n\n  /**\n   * Converts text to speech\n   * @param {string | NodeJS.ReadableStream} input - Text or stream to convert to speech\n   * @param {Object} [options] - Speech synthesis options\n   * @param {string} [options.speaker] - Voice ID to use\n   * @param {string} [options.languageCode] - Language code for the voice\n   * @param {TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig']} [options.audioConfig] - Audio configuration options\n   * @returns {Promise<NodeJS.ReadableStream>} Stream of synthesized audio. Default encoding is LINEAR16.\n   */\n  async speak(\n    input: string | NodeJS.ReadableStream,\n    options?: {\n      speaker?: string;\n      languageCode?: string;\n      audioConfig?: TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig'];\n    },\n  ): Promise<NodeJS.ReadableStream> {\n    const text = typeof input === 'string' ? input : await this.streamToString(input);\n\n    const request: TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest = {\n      input: { text },\n      voice: {\n        name: options?.speaker || this.speaker,\n        languageCode: options?.languageCode || options?.speaker?.split('-').slice(0, 2).join('-') || 'en-US',\n      },\n      audioConfig: options?.audioConfig || { audioEncoding: 'LINEAR16' },\n    };\n\n    const [response] = await this.ttsClient.synthesizeSpeech(request);\n\n    if (!response.audioContent) {\n      throw new Error('No audio content returned.');\n    }\n\n    if (typeof response.audioContent === 'string') {\n      throw new Error('Audio content is a string.');\n    }\n\n    const stream = new PassThrough();\n    stream.end(Buffer.from(response.audioContent));\n    return stream;\n  }\n\n  /**\n   * Checks if listening capabilities are enabled.\n   *\n   * @returns {Promise<{ enabled: boolean }>}\n   */\n  async getListener() {\n    return { enabled: true };\n  }\n\n  /**\n   * Converts speech to text\n   * @param {NodeJS.ReadableStream} audioStream - Audio stream to transcribe. Default encoding is LINEAR16.\n   * @param {Object} [options] - Recognition options\n   * @param {SpeechTypes.cloud.speech.v1.IRecognitionConfig} [options.config] - Recognition configuration\n   * @returns {Promise<string>} Transcribed text\n   */\n  async listen(\n    audioStream: NodeJS.ReadableStream,\n    options?: { stream?: boolean; config?: SpeechTypes.cloud.speech.v1.IRecognitionConfig },\n  ): Promise<string> {\n    const chunks: Buffer[] = [];\n    for await (const chunk of audioStream) {\n      if (typeof chunk === 'string') {\n        chunks.push(Buffer.from(chunk));\n      } else {\n        chunks.push(chunk);\n      }\n    }\n    const buffer = Buffer.concat(chunks);\n\n    let request = {\n      config: {\n        encoding: 'LINEAR16',\n        languageCode: 'en-US',\n        ...options?.config,\n      },\n      audio: {\n        content: buffer.toString('base64'),\n      },\n    };\n    const [response] = await this.speechClient.recognize(request as SpeechTypes.cloud.speech.v1.IRecognizeRequest);\n\n    if (!response.results || response.results.length === 0) {\n      throw new Error('No transcription results returned');\n    }\n\n    const transcription = response.results\n      .map((result: any) => {\n        if (!result.alternatives || result.alternatives.length === 0) {\n          return '';\n        }\n        return result.alternatives[0].transcript || '';\n      })\n      .filter((text: string) => text.length > 0)\n      .join(' ');\n\n    if (!transcription) {\n      throw new Error('No valid transcription found in results');\n    }\n\n    return transcription;\n  }\n}\n"]}

package/dist/index.d.ts

@@ -5,31 +5,115 @@ import { MastraVoice } from '@mastra/core/voice';
  * Configuration for Google Cloud Voice models
  * @interface GoogleModelConfig
  * @property {string} [apiKey] - Optional Google Cloud API key. If not provided, will use GOOGLE_API_KEY environment variable
+ * @property {string} [keyFilename] - Optional path to a service account key file. If not provided, will use GOOGLE_APPLICATION_CREDENTIALS environment variable
+ * @property {{ client_email?: string; private_key?: string }} [credentials] - Optional in-memory service account credentials
  */
 export interface GoogleModelConfig {
     apiKey?: string;
+    keyFilename?: string;
+    credentials?: {
+        client_email?: string;
+        private_key?: string;
+        [key: string]: unknown;
+    };
 }
 /**
- * GoogleVoice class provides Text-to-Speech and Speech-to-Text capabilities using Google Cloud services
+ * Configuration options for GoogleVoice
+ * @interface GoogleVoiceConfig
+ */
+export interface GoogleVoiceConfig {
+    /** Configuration for speech synthesis (TTS) */
+    speechModel?: GoogleModelConfig;
+    /** Configuration for speech recognition (STT) */
+    listeningModel?: GoogleModelConfig;
+    /** Default voice ID to use for speech synthesis */
+    speaker?: string;
+    /**
+     * Enable Vertex AI mode for enterprise deployments.
+     * When true, uses Google Cloud project-based authentication instead of API keys.
+     * Requires `project` to be set or GOOGLE_CLOUD_PROJECT environment variable.
+     */
+    vertexAI?: boolean;
+    /**
+     * Google Cloud project ID (required when vertexAI is true).
+     * Falls back to GOOGLE_CLOUD_PROJECT environment variable.
+     */
+    project?: string;
+    /**
+     * Google Cloud region for Vertex AI endpoints.
+     * Falls back to GOOGLE_CLOUD_LOCATION environment variable.
+     * @default 'us-central1'
+     */
+    location?: string;
+}
+/**
+ * GoogleVoice class provides Text-to-Speech and Speech-to-Text capabilities using Google Cloud services.
+ * Supports both standard Google Cloud API authentication and Vertex AI mode for enterprise deployments.
+ *
  * @class GoogleVoice
  * @extends MastraVoice
+ *
+ * @example Standard usage with API key
+ * ```typescript
+ * const voice = new GoogleVoice({
+ *   speechModel: { apiKey: 'your-api-key' },
+ *   speaker: 'en-US-Studio-O',
+ * });
+ * ```
+ *
+ * @example Vertex AI mode (recommended for production)
+ * ```typescript
+ * const voice = new GoogleVoice({
+ *   vertexAI: true,
+ *   project: 'your-gcp-project',
+ *   location: 'us-central1',
+ *   speaker: 'en-US-Studio-O',
+ * });
+ * ```
+ *
+ * @example Vertex AI with service account
+ * ```typescript
+ * const voice = new GoogleVoice({
+ *   vertexAI: true,
+ *   project: 'your-gcp-project',
+ *   speechModel: {
+ *     keyFilename: '/path/to/service-account.json',
+ *   },
+ * });
+ * ```
  */
 export declare class GoogleVoice extends MastraVoice {
     private ttsClient;
     private speechClient;
+    private readonly vertexAI;
+    private readonly project?;
+    private readonly location;
     /**
      * Creates an instance of GoogleVoice
-     * @param {Object} config - Configuration options
+     * @param {GoogleVoiceConfig} config - Configuration options
      * @param {GoogleModelConfig} [config.speechModel] - Configuration for speech synthesis
      * @param {GoogleModelConfig} [config.listeningModel] - Configuration for speech recognition
      * @param {string} [config.speaker] - Default voice ID to use for speech synthesis
-     * @throws {Error} If no API key is provided via config or environment variable
+     * @param {boolean} [config.vertexAI] - Enable Vertex AI mode
+     * @param {string} [config.project] - Google Cloud project ID (required for Vertex AI)
+     * @param {string} [config.location] - Google Cloud region (default: 'us-central1')
      */
-    constructor({ listeningModel, speechModel, speaker, }?: {
-        listeningModel?: GoogleModelConfig;
-        speechModel?: GoogleModelConfig;
-        speaker?: string;
-    });
+    constructor({ listeningModel, speechModel, speaker, vertexAI, project, location }?: GoogleVoiceConfig);
+    /**
+     * Check if Vertex AI mode is enabled
+     * @returns {boolean} True if using Vertex AI
+     */
+    isUsingVertexAI(): boolean;
+    /**
+     * Get the configured Google Cloud project ID
+     * @returns {string | undefined} The project ID or undefined if not set
+     */
+    getProject(): string | undefined;
+    /**
+     * Get the configured Google Cloud location/region
+     * @returns {string} The location (default: 'us-central1')
+     */
+    getLocation(): string;
     /**
      * Gets a list of available voices
      * @returns {Promise<Array<{voiceId: string, languageCodes: string[]}>>} List of available voices and their supported languages. Default language is en-US.

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,MAAM,IAAI,WAAW,EAAE,MAAM,0CAA0C,CAAC;AAEtF,OAAO,KAAK,EAAE,MAAM,IAAI,iBAAiB,EAAE,MAAM,kDAAkD,CAAC;AACpG,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AAEjD;;;;GAIG;AACH,MAAM,WAAW,iBAAiB;IAChC,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAID;;;;GAIG;AACH,qBAAa,WAAY,SAAQ,WAAW;IAC1C,OAAO,CAAC,SAAS,CAAqB;IACtC,OAAO,CAAC,YAAY,CAAe;IAEnC;;;;;;;OAOG;gBACS,EACV,cAAc,EACd,WAAW,EACX,OAAO,GACR,GAAE;QACD,cAAc,CAAC,EAAE,iBAAiB,CAAC;QACnC,WAAW,CAAC,EAAE,iBAAiB,CAAC;QAChC,OAAO,CAAC,EAAE,MAAM,CAAC;KACb;IAgCN;;;OAGG;IACG,WAAW,CAAC,EAAE,YAAsB,EAAE,GAAE;QAAE,YAAY,CAAC,EAAE,MAAM,CAAA;KAAO;;;;YAY9D,cAAc;IAY5B;;;;;;;;OAQG;IACG,KAAK,CACT,KAAK,EAAE,MAAM,GAAG,MAAM,CAAC,cAAc,EACrC,OAAO,CAAC,EAAE;QACR,OAAO,CAAC,EAAE,MAAM,CAAC;QACjB,YAAY,CAAC,EAAE,MAAM,CAAC;QACtB,WAAW,CAAC,EAAE,iBAAiB,CAAC,KAAK,CAAC,YAAY,CAAC,EAAE,CAAC,wBAAwB,CAAC,aAAa,CAAC,CAAC;KAC/F,GACA,OAAO,CAAC,MAAM,CAAC,cAAc,CAAC;IA6BjC;;;;OAIG;IACG,WAAW;;;IAIjB;;;;;;OAMG;IACG,MAAM,CACV,WAAW,EAAE,MAAM,CAAC,cAAc,EAClC,OAAO,CAAC,EAAE;QAAE,MAAM,CAAC,EAAE,OAAO,CAAC;QAAC,MAAM,CAAC,EAAE,WAAW,CAAC,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,kBAAkB,CAAA;KAAE,GACtF,OAAO,CAAC,MAAM,CAAC;CA+CnB"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,MAAM,IAAI,WAAW,EAAE,MAAM,0CAA0C,CAAC;AAEtF,OAAO,KAAK,EAAE,MAAM,IAAI,iBAAiB,EAAE,MAAM,kDAAkD,CAAC;AACpG,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AAEjD;;;;;;GAMG;AACH,MAAM,WAAW,iBAAiB;IAChC,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,WAAW,CAAC,EAAE;QACZ,YAAY,CAAC,EAAE,MAAM,CAAC;QACtB,WAAW,CAAC,EAAE,MAAM,CAAC;QACrB,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC;KACxB,CAAC;CACH;AAED;;;GAGG;AACH,MAAM,WAAW,iBAAiB;IAChC,+CAA+C;IAC/C,WAAW,CAAC,EAAE,iBAAiB,CAAC;IAChC,iDAAiD;IACjD,cAAc,CAAC,EAAE,iBAAiB,CAAC;IACnC,mDAAmD;IACnD,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB;;;;OAIG;IACH,QAAQ,CAAC,EAAE,OAAO,CAAC;IACnB;;;OAGG;IACH,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB;;;;OAIG;IACH,QAAQ,CAAC,EAAE,MAAM,CAAC;CACnB;AAuED;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAmCG;AACH,qBAAa,WAAY,SAAQ,WAAW;IAC1C,OAAO,CAAC,SAAS,CAAqB;IACtC,OAAO,CAAC,YAAY,CAAe;IACnC,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAAU;IACnC,OAAO,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAS;IAClC,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAAS;IAElC;;;;;;;;;OASG;gBACS,EAAE,cAAc,EAAE,WAAW,EAAE,OAAO,EAAE,QAAgB,EAAE,OAAO,EAAE,QAAQ,EAAE,GAAE,iBAAsB;IAoDjH;;;OAGG;IACH,eAAe,IAAI,OAAO;IAI1B;;;OAGG;IACH,UAAU,IAAI,MAAM,GAAG,SAAS;IAIhC;;;OAGG;IACH,WAAW,IAAI,MAAM;IAIrB;;;OAGG;IACG,WAAW,CAAC,EAAE,YAAsB,EAAE,GAAE;QAAE,YAAY,CAAC,EAAE,MAAM,CAAA;KAAO;;;;YAU9D,cAAc;IAY5B;;;;;;;;OAQG;IACG,KAAK,CACT,KAAK,EAAE,MAAM,GAAG,MAAM,CAAC,cAAc,EACrC,OAAO,CAAC,EAAE;QACR,OAAO,CAAC,EAAE,MAAM,CAAC;QACjB,YAAY,CAAC,EAAE,MAAM,CAAC;QACtB,WAAW,CAAC,EAAE,iBAAiB,CAAC,KAAK,CAAC,YAAY,CAAC,EAAE,CAAC,wBAAwB,CAAC,aAAa,CAAC,CAAC;KAC/F,GACA,OAAO,CAAC,MAAM,CAAC,cAAc,CAAC;IA2BjC;;;;OAIG;IACG,WAAW;;;IAIjB;;;;;;OAMG;IACG,MAAM,CACV,WAAW,EAAE,MAAM,CAAC,cAAc,EAClC,OAAO,CAAC,EAAE;QAAE,MAAM,CAAC,EAAE,OAAO,CAAC;QAAC,MAAM,CAAC,EAAE,WAAW,CAAC,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,kBAAkB,CAAA;KAAE,GACtF,OAAO,CAAC,MAAM,CAAC;CA2CnB"}