@mastra/voice-google 0.0.0-vnext-inngest-20250508122351 → 0.0.0-vnext-20251104230439

package/dist/index.js

@@ -4,6 +4,34 @@ import { TextToSpeechClient } from '@google-cloud/text-to-speech';
 import { MastraVoice } from '@mastra/core/voice';
 
 // src/index.ts
+var resolveAuthConfig = (modelConfig, fallback) => {
+  const resolved = {};
+  const apiKey = modelConfig?.apiKey ?? fallback.apiKey;
+  if (apiKey) {
+    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) => {
+  if (config.credentials) {
+    return { credentials: config.credentials };
+  }
+  if (config.keyFilename) {
+    return { keyFilename: config.keyFilename };
+  }
+  if (config.apiKey) {
+    return { apiKey: config.apiKey };
+  }
+  return {};
+};
 var DEFAULT_VOICE = "en-US-Casual-K";
 var GoogleVoice = class extends MastraVoice {
   ttsClient;
@@ -14,7 +42,6 @@ var GoogleVoice = class extends MastraVoice {
    * @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
    */
   constructor({
     listeningModel,
@@ -22,43 +49,41 @@ var GoogleVoice = class extends MastraVoice {
     speaker
   } = {}) {
     const defaultApiKey = process.env.GOOGLE_API_KEY;
+    const defaultKeyFilename = process.env.GOOGLE_APPLICATION_CREDENTIALS;
     const defaultSpeaker = DEFAULT_VOICE;
+    const sharedFallback = {
+      apiKey: defaultApiKey ?? speechModel?.apiKey ?? listeningModel?.apiKey,
+      keyFilename: defaultKeyFilename ?? speechModel?.keyFilename ?? listeningModel?.keyFilename,
+      credentials: speechModel?.credentials ?? listeningModel?.credentials
+    };
+    const speechAuthConfig = resolveAuthConfig(speechModel, sharedFallback);
+    const listeningAuthConfig = resolveAuthConfig(listeningModel, sharedFallback);
     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 TextToSpeechClient({
-      apiKey: this.speechModel?.apiKey || defaultApiKey
-    });
-    this.speechClient = new SpeechClient({
-      apiKey: this.listeningModel?.apiKey || defaultApiKey
-    });
+    const ttsOptions = buildAuthOptions(speechAuthConfig);
+    const speechOptions = buildAuthOptions(listeningAuthConfig);
+    this.ttsClient = new TextToSpeechClient(ttsOptions);
+    this.speechClient = new SpeechClient(speechOptions);
   }
   /**
    * 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 = [];
@@ -81,27 +106,33 @@ var GoogleVoice = class extends 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 = new PassThrough();
-      stream.end(Buffer.from(response.audioContent));
-      return stream;
-    }, "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 = new PassThrough();
+    stream.end(Buffer.from(response.audioContent));
+    return stream;
+  }
+  /**
+   * Checks if listening capabilities are enabled.
+   *
+   * @returns {Promise<{ enabled: boolean }>}
+   */
+  async getListener() {
+    return { enabled: true };
   }
   /**
    * Converts speech to text
@@ -111,44 +142,42 @@ var GoogleVoice = class extends 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;
   }
 };
 
 export { GoogleVoice };
+//# sourceMappingURL=index.js.map
+//# sourceMappingURL=index.js.map

package/dist/index.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/index.ts"],"names":[],"mappings":";;;;;;AA6BA,IAAM,iBAAA,GAAoB,CAAC,WAAA,EAA4C,QAAA,KAAqC;AAC1G,EAAA,MAAM,WAAuB,EAAC;AAE9B,EAAA,MAAM,MAAA,GAAS,WAAA,EAAa,MAAA,IAAU,QAAA,CAAS,MAAA;AAC/C,EAAA,IAAI,MAAA,EAAQ;AACV,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,CAAC,MAAA,KAA4C;AACpE,EAAA,IAAI,OAAO,WAAA,EAAa;AACtB,IAAA,OAAO,EAAE,WAAA,EAAa,MAAA,CAAO,WAAA,EAAY;AAAA,EAC3C;AAEA,EAAA,IAAI,OAAO,WAAA,EAAa;AACtB,IAAA,OAAO,EAAE,WAAA,EAAa,MAAA,CAAO,WAAA,EAAY;AAAA,EAC3C;AAEA,EAAA,IAAI,OAAO,MAAA,EAAQ;AACjB,IAAA,OAAO,EAAE,MAAA,EAAQ,MAAA,CAAO,MAAA,EAAO;AAAA,EACjC;AAEA,EAAA,OAAO,EAAC;AACV,CAAA;AAEA,IAAM,aAAA,GAAgB,gBAAA;AAOf,IAAM,WAAA,GAAN,cAA0B,WAAA,CAAY;AAAA,EACnC,SAAA;AAAA,EACA,YAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EASR,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,kBAAA,GAAqB,QAAQ,GAAA,CAAI,8BAAA;AACvC,IAAA,MAAM,cAAA,GAAiB,aAAA;AAEvB,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;AAAA,KAC3D;AAEA,IAAA,MAAM,gBAAA,GAAmB,iBAAA,CAAkB,WAAA,EAAa,cAAc,CAAA;AACtE,IAAA,MAAM,mBAAA,GAAsB,iBAAA,CAAkB,cAAA,EAAgB,cAAc,CAAA;AAE5E,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,MAAM,UAAA,GAAa,iBAAiB,gBAAgB,CAAA;AACpD,IAAA,MAAM,aAAA,GAAgB,iBAAiB,mBAAmB,CAAA;AAE1D,IAAA,IAAA,CAAK,SAAA,GAAY,IAAI,kBAAA,CAAmB,UAAU,CAAA;AAElD,IAAA,IAAA,CAAK,YAAA,GAAe,IAAI,YAAA,CAAa,aAAa,CAAA;AAAA,EACpD;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,MAAM,MAAA,GAAS,IAAI,WAAA,EAAY;AAC/B,IAAA,MAAA,CAAO,GAAA,CAAI,MAAA,CAAO,IAAA,CAAK,QAAA,CAAS,YAAY,CAAC,CAAA;AAC7C,IAAA,OAAO,MAAA;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.js","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 * @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\ntype AuthConfig = Pick<GoogleModelConfig, 'apiKey' | 'keyFilename' | 'credentials'>;\n\ntype GoogleClientOptions = AuthConfig;\n\nconst resolveAuthConfig = (modelConfig: GoogleModelConfig | undefined, fallback: AuthConfig): AuthConfig => {\n  const resolved: AuthConfig = {};\n\n  const apiKey = modelConfig?.apiKey ?? fallback.apiKey;\n  if (apiKey) {\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 = (config: AuthConfig): GoogleClientOptions => {\n  if (config.credentials) {\n    return { credentials: config.credentials };\n  }\n\n  if (config.keyFilename) {\n    return { keyFilename: config.keyFilename };\n  }\n\n  if (config.apiKey) {\n    return { apiKey: config.apiKey };\n  }\n\n  return {};\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   */\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 defaultKeyFilename = process.env.GOOGLE_APPLICATION_CREDENTIALS;\n    const defaultSpeaker = DEFAULT_VOICE;\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    };\n\n    const speechAuthConfig = resolveAuthConfig(speechModel, sharedFallback);\n    const listeningAuthConfig = resolveAuthConfig(listeningModel, sharedFallback);\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    const ttsOptions = buildAuthOptions(speechAuthConfig);\n    const speechOptions = buildAuthOptions(listeningAuthConfig);\n\n    this.ttsClient = new TextToSpeechClient(ttsOptions);\n\n    this.speechClient = new SpeechClient(speechOptions);\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/package.json

@@ -1,10 +1,11 @@
 {
   "name": "@mastra/voice-google",
-  "version": "0.0.0-vnext-inngest-20250508122351",
+  "version": "0.0.0-vnext-20251104230439",
   "description": "Mastra Google voice integration",
   "type": "module",
   "files": [
-    "dist"
+    "dist",
+    "CHANGELOG.md"
   ],
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
@@ -15,30 +16,43 @@
         "default": "./dist/index.js"
       },
       "require": {
-        "types": "./dist/index.d.cts",
+        "types": "./dist/index.d.ts",
         "default": "./dist/index.cjs"
       }
     },
     "./package.json": "./package.json"
   },
-  "license": "Elastic-2.0",
+  "license": "Apache-2.0",
   "dependencies": {
     "@google-cloud/speech": "^6.7.1",
-    "@google-cloud/text-to-speech": "^6.0.1",
-    "zod": "^3.24.3",
-    "@mastra/core": "0.0.0-vnext-inngest-20250508122351"
+    "@google-cloud/text-to-speech": "^6.3.1"
   },
   "devDependencies": {
-    "@types/node": "^20.17.32",
-    "eslint": "^9.25.1",
-    "tsup": "^8.4.0",
+    "@types/node": "^20.19.0",
+    "eslint": "^9.37.0",
+    "tsup": "^8.5.0",
     "typescript": "^5.8.3",
-    "vitest": "^2.1.9",
-    "@internal/lint": "0.0.0-vnext-inngest-20250508122351"
+    "vitest": "^3.2.4",
+    "@internal/lint": "0.0.0-vnext-20251104230439",
+    "@internal/types-builder": "0.0.0-vnext-20251104230439",
+    "@mastra/core": "0.0.0-vnext-20251104230439"
+  },
+  "peerDependencies": {
+    "zod": "^3.25.0 || ^4.0.0",
+    "@mastra/core": "0.0.0-vnext-20251104230439"
+  },
+  "homepage": "https://mastra.ai",
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/mastra-ai/mastra.git",
+    "directory": "voice/google"
+  },
+  "bugs": {
+    "url": "https://github.com/mastra-ai/mastra/issues"
   },
   "scripts": {
-    "build": "tsup src/index.ts --format esm,cjs --experimental-dts --clean --treeshake=smallest --splitting",
-    "build:watch": "pnpm build --watch",
+    "build": "tsup --silent --config tsup.config.ts",
+    "build:watch": "tsup --watch --silent --config tsup.config.ts",
     "test": "vitest run",
     "lint": "eslint ."
   }

package/dist/_tsup-dts-rollup.d.cts

@@ -1,73 +0,0 @@
-import type { google } from '@google-cloud/text-to-speech/build/protos/protos';
-import type { google as google_2 } from '@google-cloud/speech/build/protos/protos';
-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
- */
-export declare interface GoogleModelConfig {
-    apiKey?: string;
-}
-
-/**
- * GoogleVoice class provides Text-to-Speech and Speech-to-Text capabilities using Google Cloud services
- * @class GoogleVoice
- * @extends MastraVoice
- */
-export declare class GoogleVoice extends MastraVoice {
-    private ttsClient;
-    private speechClient;
-    /**
-     * Creates an instance of GoogleVoice
-     * @param {Object} 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
-     */
-    constructor({ listeningModel, speechModel, speaker, }?: {
-        listeningModel?: GoogleModelConfig;
-        speechModel?: GoogleModelConfig;
-        speaker?: 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.
-     */
-    getSpeakers({ languageCode }?: {
-        languageCode?: string;
-    }): Promise<{
-        voiceId: string;
-        languageCodes: string[];
-    }[]>;
-    private streamToString;
-    /**
-     * Converts text to speech
-     * @param {string | NodeJS.ReadableStream} input - Text or stream to convert to speech
-     * @param {Object} [options] - Speech synthesis options
-     * @param {string} [options.speaker] - Voice ID to use
-     * @param {string} [options.languageCode] - Language code for the voice
-     * @param {TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig']} [options.audioConfig] - Audio configuration options
-     * @returns {Promise<NodeJS.ReadableStream>} Stream of synthesized audio. Default encoding is LINEAR16.
-     */
-    speak(input: string | NodeJS.ReadableStream, options?: {
-        speaker?: string;
-        languageCode?: string;
-        audioConfig?: google.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig'];
-    }): Promise<NodeJS.ReadableStream>;
-    /**
-     * Converts speech to text
-     * @param {NodeJS.ReadableStream} audioStream - Audio stream to transcribe. Default encoding is LINEAR16.
-     * @param {Object} [options] - Recognition options
-     * @param {SpeechTypes.cloud.speech.v1.IRecognitionConfig} [options.config] - Recognition configuration
-     * @returns {Promise<string>} Transcribed text
-     */
-    listen(audioStream: NodeJS.ReadableStream, options?: {
-        stream?: boolean;
-        config?: google_2.cloud.speech.v1.IRecognitionConfig;
-    }): Promise<string>;
-}
-
-export { }

package/dist/_tsup-dts-rollup.d.ts

@@ -1,73 +0,0 @@
-import type { google } from '@google-cloud/text-to-speech/build/protos/protos';
-import type { google as google_2 } from '@google-cloud/speech/build/protos/protos';
-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
- */
-export declare interface GoogleModelConfig {
-    apiKey?: string;
-}
-
-/**
- * GoogleVoice class provides Text-to-Speech and Speech-to-Text capabilities using Google Cloud services
- * @class GoogleVoice
- * @extends MastraVoice
- */
-export declare class GoogleVoice extends MastraVoice {
-    private ttsClient;
-    private speechClient;
-    /**
-     * Creates an instance of GoogleVoice
-     * @param {Object} 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
-     */
-    constructor({ listeningModel, speechModel, speaker, }?: {
-        listeningModel?: GoogleModelConfig;
-        speechModel?: GoogleModelConfig;
-        speaker?: 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.
-     */
-    getSpeakers({ languageCode }?: {
-        languageCode?: string;
-    }): Promise<{
-        voiceId: string;
-        languageCodes: string[];
-    }[]>;
-    private streamToString;
-    /**
-     * Converts text to speech
-     * @param {string | NodeJS.ReadableStream} input - Text or stream to convert to speech
-     * @param {Object} [options] - Speech synthesis options
-     * @param {string} [options.speaker] - Voice ID to use
-     * @param {string} [options.languageCode] - Language code for the voice
-     * @param {TextToSpeechTypes.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig']} [options.audioConfig] - Audio configuration options
-     * @returns {Promise<NodeJS.ReadableStream>} Stream of synthesized audio. Default encoding is LINEAR16.
-     */
-    speak(input: string | NodeJS.ReadableStream, options?: {
-        speaker?: string;
-        languageCode?: string;
-        audioConfig?: google.cloud.texttospeech.v1.ISynthesizeSpeechRequest['audioConfig'];
-    }): Promise<NodeJS.ReadableStream>;
-    /**
-     * Converts speech to text
-     * @param {NodeJS.ReadableStream} audioStream - Audio stream to transcribe. Default encoding is LINEAR16.
-     * @param {Object} [options] - Recognition options
-     * @param {SpeechTypes.cloud.speech.v1.IRecognitionConfig} [options.config] - Recognition configuration
-     * @returns {Promise<string>} Transcribed text
-     */
-    listen(audioStream: NodeJS.ReadableStream, options?: {
-        stream?: boolean;
-        config?: google_2.cloud.speech.v1.IRecognitionConfig;
-    }): Promise<string>;
-}
-
-export { }

package/dist/index.d.cts

@@ -1,2 +0,0 @@
-export { GoogleModelConfig } from './_tsup-dts-rollup.cjs';
-export { GoogleVoice } from './_tsup-dts-rollup.cjs';