@mastra/voice-openai 0.11.6 → 0.11.7-alpha.0

package/CHANGELOG.md

@@ -1,5 +1,16 @@
 # @mastra/voice-openai
 
+## 0.11.7-alpha.0
+
+### Patch Changes
+
+- improve options handling for speak method ([#8228](https://github.com/mastra-ai/mastra/pull/8228))
+
+- Update peer deps ([#8154](https://github.com/mastra-ai/mastra/pull/8154))
+
+- Updated dependencies [[`504438b`](https://github.com/mastra-ai/mastra/commit/504438b961bde211071186bba63a842c4e3db879), [`a7243e2`](https://github.com/mastra-ai/mastra/commit/a7243e2e58762667a6e3921e755e89d6bb0a3282), [`7fceb0a`](https://github.com/mastra-ai/mastra/commit/7fceb0a327d678e812f90f5387c5bc4f38bd039e), [`df64f9e`](https://github.com/mastra-ai/mastra/commit/df64f9ef814916fff9baedd861c988084e7c41de), [`809eea0`](https://github.com/mastra-ai/mastra/commit/809eea092fa80c3f69b9eaf078d843b57fd2a88e), [`683e5a1`](https://github.com/mastra-ai/mastra/commit/683e5a1466e48b686825b2c11f84680f296138e4), [`3679378`](https://github.com/mastra-ai/mastra/commit/3679378673350aa314741dc826f837b1984149bc), [`7775bc2`](https://github.com/mastra-ai/mastra/commit/7775bc20bb1ad1ab24797fb420e4f96c65b0d8ec), [`db1891a`](https://github.com/mastra-ai/mastra/commit/db1891a4707443720b7cd8a260dc7e1d49b3609c), [`e8f379d`](https://github.com/mastra-ai/mastra/commit/e8f379d390efa264c4e0874f9ac0cf8839b07777), [`652066b`](https://github.com/mastra-ai/mastra/commit/652066bd1efc6bb6813ba950ed1d7573e8b7d9d4), [`ea8d386`](https://github.com/mastra-ai/mastra/commit/ea8d386cd8c5593664515fd5770c06bf2aa980ef), [`c2a4919`](https://github.com/mastra-ai/mastra/commit/c2a4919ba6797d8bdb1509e02287496eef69303e), [`0130986`](https://github.com/mastra-ai/mastra/commit/0130986fc62d0edcc626dd593282661dbb9af141)]:
+  - @mastra/core@0.19.0-alpha.1
+
 ## 0.11.6
 
 ### Patch Changes

package/dist/index.cjs

@@ -111,13 +111,15 @@ var OpenAIVoice = class extends voice.MastraVoice {
     if (input.trim().length === 0) {
       throw new Error("Input text is empty");
     }
+    const { speaker, responseFormat, speed, ...otherOptions } = options || {};
     const audio = await this.traced(async () => {
       const response = await this.speechClient.audio.speech.create({
         model: this.speechModel?.name ?? "tts-1",
-        voice: options?.speaker ?? this.speaker,
-        response_format: options?.responseFormat ?? "mp3",
+        voice: speaker ?? this.speaker,
+        response_format: responseFormat ?? "mp3",
         input,
-        speed: options?.speed || 1
+        speed: speed || 1,
+        ...otherOptions
       });
       const passThrough = new stream.PassThrough();
       const buffer = Buffer.from(await response.arrayBuffer());

package/dist/index.cjs.map

@@ -1 +1 @@
-{"version":3,"sources":["../src/index.ts"],"names":["MastraVoice","OpenAI","PassThrough"],"mappings":";;;;;;;;;;;AAyBO,IAAM,WAAA,GAAN,cAA0BA,iBAAA,CAAY;AAAA,EAC3C,YAAA;AAAA,EACA,eAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAWA,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;AAAA,MACzB,IAAA,EAAM,OAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AACA,IAAA,MAAM,qBAAA,GAAwB;AAAA,MAC5B,IAAA,EAAM,WAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AAEA,IAAA,KAAA,CAAM;AAAA,MACJ,WAAA,EAAa;AAAA,QACX,IAAA,EAAM,WAAA,EAAa,IAAA,IAAQ,kBAAA,CAAmB,IAAA;AAAA,QAC9C,MAAA,EAAQ,WAAA,EAAa,MAAA,IAAU,kBAAA,CAAmB;AAAA,OACpD;AAAA,MACA,cAAA,EAAgB;AAAA,QACd,IAAA,EAAM,cAAA,EAAgB,IAAA,IAAQ,qBAAA,CAAsB,IAAA;AAAA,QACpD,MAAA,EAAQ,cAAA,EAAgB,MAAA,IAAU,qBAAA,CAAsB;AAAA,OAC1D;AAAA,MACA,SAAS,OAAA,IAAW;AAAA,KACrB,CAAA;AAED,IAAA,MAAM,YAAA,GAAe,aAAa,MAAA,IAAU,aAAA;AAC5C,IAAA,IAAI,CAAC,YAAA,EAAc;AACjB,MAAA,MAAM,IAAI,MAAM,sCAAsC,CAAA;AAAA,IACxD;AACA,IAAA,IAAA,CAAK,eAAe,IAAIC,uBAAA,CAAO,EAAE,MAAA,EAAQ,cAAc,CAAA;AAEvD,IAAA,MAAM,eAAA,GAAkB,gBAAgB,MAAA,IAAU,aAAA;AAClD,IAAA,IAAI,CAAC,eAAA,EAAiB;AACpB,MAAA,MAAM,IAAI,MAAM,yCAAyC,CAAA;AAAA,IAC3D;AACA,IAAA,IAAA,CAAK,kBAAkB,IAAIA,uBAAA,CAAO,EAAE,MAAA,EAAQ,iBAAiB,CAAA;AAE7D,IAAA,IAAI,CAAC,IAAA,CAAK,YAAA,IAAgB,CAAC,KAAK,eAAA,EAAiB;AAC/C,MAAA,MAAM,IAAI,MAAM,0FAA0F,CAAA;AAAA,IAC5G;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EASA,MAAM,WAAA,GAA0D;AAC9D,IAAA,IAAI,CAAC,KAAK,WAAA,EAAa;AACrB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,OAAO;AAAA,MACL,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,SAAA,EAAU;AAAA,MACrB,EAAE,SAAS,KAAA,EAAM;AAAA,MACjB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA;AAAO,KACpB;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,KAAA,CACJ,KAAA,EACA,OAAA,EAKgC;AAChC,IAAA,IAAI,CAAC,KAAK,YAAA,EAAc;AACtB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,MAAA,MAAM,SAAmB,EAAC;AAC1B,MAAA,WAAA,MAAiB,SAAS,KAAA,EAAO;AAC/B,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,KAAA,GAAQ,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA,CAAE,SAAS,OAAO,CAAA;AAAA,IAChD;AAEA,IAAA,IAAI,KAAA,CAAM,IAAA,EAAK,CAAE,MAAA,KAAW,CAAA,EAAG;AAC7B,MAAA,MAAM,IAAI,MAAM,qBAAqB,CAAA;AAAA,IACvC;AAEA,IAAA,MAAM,KAAA,GAAQ,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AAC1C,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,YAAA,CAAc,KAAA,CAAM,OAAO,MAAA,CAAO;AAAA,QAC5D,KAAA,EAAO,IAAA,CAAK,WAAA,EAAa,IAAA,IAAQ,OAAA;AAAA,QACjC,KAAA,EAAQ,OAAA,EAAS,OAAA,IAAW,IAAA,CAAK,OAAA;AAAA,QACjC,eAAA,EAAiB,SAAS,cAAA,IAAkB,KAAA;AAAA,QAC5C,KAAA;AAAA,QACA,KAAA,EAAO,SAAS,KAAA,IAAS;AAAA,OAC1B,CAAA;AAED,MAAA,MAAM,WAAA,GAAc,IAAIC,kBAAA,EAAY;AACpC,MAAA,MAAM,SAAS,MAAA,CAAO,IAAA,CAAK,MAAM,QAAA,CAAS,aAAa,CAAA;AACvD,MAAA,WAAA,CAAY,IAAI,MAAM,CAAA;AACtB,MAAA,OAAO,WAAA;AAAA,IACT,CAAA,EAAG,oBAAoB,CAAA,EAAE;AAEzB,IAAA,OAAO,KAAA;AAAA,EACT;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAOA,MAAM,WAAA,GAAc;AAClB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,OAAO,EAAE,SAAS,KAAA,EAAM;AAAA,IAC1B;AACA,IAAA,OAAO,EAAE,SAAS,IAAA,EAAK;AAAA,EACzB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,MAAA,CACJ,WAAA,EACA,OAAA,EAIiB;AACjB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,MAAM,IAAI,MAAM,gCAAgC,CAAA;AAAA,IAClD;AAEA,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,WAAA,GAAc,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA;AAExC,IAAA,MAAM,IAAA,GAAO,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AACzC,MAAA,MAAM,EAAE,QAAA,EAAU,GAAG,YAAA,EAAa,GAAI,WAAW,EAAC;AAClD,MAAA,MAAM,IAAA,GAAO,IAAI,IAAA,CAAK,CAAC,WAAW,CAAA,EAAG,CAAA,MAAA,EAAS,QAAA,IAAY,KAAK,CAAA,CAAE,CAAA;AAEjE,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,eAAA,CAAiB,KAAA,CAAM,eAAe,MAAA,CAAO;AAAA,QACvE,KAAA,EAAO,IAAA,CAAK,cAAA,EAAgB,IAAA,IAAQ,WAAA;AAAA,QACpC,IAAA;AAAA,QACA,GAAG;AAAA,OACJ,CAAA;AAED,MAAA,OAAO,QAAA,CAAS,IAAA;AAAA,IAClB,CAAA,EAAG,qBAAqB,CAAA,EAAE;AAE1B,IAAA,OAAO,IAAA;AAAA,EACT;AACF","file":"index.cjs","sourcesContent":["import { PassThrough } from 'stream';\n\nimport { MastraVoice } from '@mastra/core/voice';\nimport OpenAI from 'openai';\n\ntype OpenAIVoiceId = 'alloy' | 'echo' | 'fable' | 'onyx' | 'nova' | 'shimmer' | 'ash' | 'coral' | 'sage';\ntype OpenAIModel = 'tts-1' | 'tts-1-hd' | 'whisper-1';\n\nexport interface OpenAIConfig {\n  name?: OpenAIModel;\n  apiKey?: string;\n}\n\nexport interface OpenAIVoiceConfig {\n  speech?: {\n    model: 'tts-1' | 'tts-1-hd';\n    apiKey?: string;\n    speaker?: OpenAIVoiceId;\n  };\n  listening?: {\n    model: 'whisper-1';\n    apiKey?: string;\n  };\n}\n\nexport class OpenAIVoice extends MastraVoice {\n  speechClient?: OpenAI;\n  listeningClient?: OpenAI;\n\n  /**\n   * Constructs an instance of OpenAIVoice with optional configurations for speech and listening models.\n   *\n   * @param {Object} [config] - Configuration options for the OpenAIVoice instance.\n   * @param {OpenAIConfig} [config.listeningModel] - Configuration for the listening model, including model name and API key.\n   * @param {OpenAIConfig} [config.speechModel] - Configuration for the speech model, including model name and API key.\n   * @param {string} [config.speaker] - The default speaker's voice to use for speech synthesis.\n   * @throws {Error} - Throws an error if no API key is provided for either the speech or listening model.\n   */\n  constructor({\n    listeningModel,\n    speechModel,\n    speaker,\n  }: {\n    listeningModel?: OpenAIConfig;\n    speechModel?: OpenAIConfig;\n    speaker?: string;\n  } = {}) {\n    const defaultApiKey = process.env.OPENAI_API_KEY;\n    const defaultSpeechModel = {\n      name: 'tts-1',\n      apiKey: defaultApiKey,\n    };\n    const defaultListeningModel = {\n      name: 'whisper-1',\n      apiKey: defaultApiKey,\n    };\n\n    super({\n      speechModel: {\n        name: speechModel?.name ?? defaultSpeechModel.name,\n        apiKey: speechModel?.apiKey ?? defaultSpeechModel.apiKey,\n      },\n      listeningModel: {\n        name: listeningModel?.name ?? defaultListeningModel.name,\n        apiKey: listeningModel?.apiKey ?? defaultListeningModel.apiKey,\n      },\n      speaker: speaker ?? 'alloy',\n    });\n\n    const speechApiKey = speechModel?.apiKey || defaultApiKey;\n    if (!speechApiKey) {\n      throw new Error('No API key provided for speech model');\n    }\n    this.speechClient = new OpenAI({ apiKey: speechApiKey });\n\n    const listeningApiKey = listeningModel?.apiKey || defaultApiKey;\n    if (!listeningApiKey) {\n      throw new Error('No API key provided for listening model');\n    }\n    this.listeningClient = new OpenAI({ apiKey: listeningApiKey });\n\n    if (!this.speechClient && !this.listeningClient) {\n      throw new Error('At least one of OPENAI_API_KEY, speechModel.apiKey, or listeningModel.apiKey must be set');\n    }\n  }\n\n  /**\n   * Retrieves a list of available speakers for the speech model.\n   *\n   * @returns {Promise<Array<{ voiceId: OpenAIVoiceId }>>} - A promise that resolves to an array of objects,\n   * each containing a `voiceId` representing an available speaker.\n   * @throws {Error} - Throws an error if the speech model is not configured.\n   */\n  async getSpeakers(): Promise<Array<{ voiceId: OpenAIVoiceId }>> {\n    if (!this.speechModel) {\n      throw new Error('Speech model not configured');\n    }\n\n    return [\n      { voiceId: 'alloy' },\n      { voiceId: 'echo' },\n      { voiceId: 'fable' },\n      { voiceId: 'onyx' },\n      { voiceId: 'nova' },\n      { voiceId: 'shimmer' },\n      { voiceId: 'ash' },\n      { voiceId: 'coral' },\n      { voiceId: 'sage' },\n    ];\n  }\n\n  /**\n   * Converts text or audio input into speech using the configured speech model.\n   *\n   * @param {string | NodeJS.ReadableStream} input - The text or audio stream to be converted into speech.\n   * @param {Object} [options] - Optional parameters for the speech synthesis.\n   * @param {string} [options.speaker] - The speaker's voice to use for the speech synthesis.\n   * @param {number} [options.speed] - The speed at which the speech should be synthesized.\n   * @returns {Promise<NodeJS.ReadableStream>} - A promise that resolves to a readable stream of the synthesized audio.\n   * @throws {Error} - Throws an error if the speech model is not configured or if the input text is empty.\n   */\n  async speak(\n    input: string | NodeJS.ReadableStream,\n    options?: {\n      speaker?: string;\n      speed?: number;\n      [key: string]: any;\n    },\n  ): Promise<NodeJS.ReadableStream> {\n    if (!this.speechClient) {\n      throw new Error('Speech model not configured');\n    }\n\n    if (typeof input !== 'string') {\n      const chunks: Buffer[] = [];\n      for await (const chunk of input) {\n        if (typeof chunk === 'string') {\n          chunks.push(Buffer.from(chunk));\n        } else {\n          chunks.push(chunk);\n        }\n      }\n      input = Buffer.concat(chunks).toString('utf-8');\n    }\n\n    if (input.trim().length === 0) {\n      throw new Error('Input text is empty');\n    }\n\n    const audio = await this.traced(async () => {\n      const response = await this.speechClient!.audio.speech.create({\n        model: this.speechModel?.name ?? 'tts-1',\n        voice: (options?.speaker ?? this.speaker) as OpenAIVoiceId,\n        response_format: options?.responseFormat ?? 'mp3',\n        input,\n        speed: options?.speed || 1.0,\n      });\n\n      const passThrough = new PassThrough();\n      const buffer = Buffer.from(await response.arrayBuffer());\n      passThrough.end(buffer);\n      return passThrough;\n    }, 'voice.openai.speak')();\n\n    return audio;\n  }\n\n  /**\n   * Checks if listening capabilities are enabled.\n   *\n   * @returns {Promise<{ enabled: boolean }>}\n   */\n  async getListener() {\n    if (!this.listeningClient) {\n      return { enabled: false };\n    }\n    return { enabled: true };\n  }\n\n  /**\n   * Transcribes audio from a given stream using the configured listening model.\n   *\n   * @param {NodeJS.ReadableStream} audioStream - The audio stream to be transcribed.\n   * @param {Object} [options] - Optional parameters for the transcription.\n   * @param {string} [options.filetype] - The file type of the audio stream.\n   *                                      Supported types include 'mp3', 'mp4', 'mpeg', 'mpga', 'm4a', 'wav', 'webm'.\n   * @returns {Promise<string>} - A promise that resolves to the transcribed text.\n   * @throws {Error} - Throws an error if the listening model is not configured.\n   */\n  async listen(\n    audioStream: NodeJS.ReadableStream,\n    options?: {\n      filetype?: 'mp3' | 'mp4' | 'mpeg' | 'mpga' | 'm4a' | 'wav' | 'webm';\n      [key: string]: any;\n    },\n  ): Promise<string> {\n    if (!this.listeningClient) {\n      throw new Error('Listening model not configured');\n    }\n\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 audioBuffer = Buffer.concat(chunks);\n\n    const text = await this.traced(async () => {\n      const { filetype, ...otherOptions } = options || {};\n      const file = new File([audioBuffer], `audio.${filetype || 'mp3'}`);\n\n      const response = await this.listeningClient!.audio.transcriptions.create({\n        model: this.listeningModel?.name || 'whisper-1',\n        file: file as any,\n        ...otherOptions,\n      });\n\n      return response.text;\n    }, 'voice.openai.listen')();\n\n    return text;\n  }\n}\n"]}
+{"version":3,"sources":["../src/index.ts"],"names":["MastraVoice","OpenAI","PassThrough"],"mappings":";;;;;;;;;;;AAyBO,IAAM,WAAA,GAAN,cAA0BA,iBAAA,CAAY;AAAA,EAC3C,YAAA;AAAA,EACA,eAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAWA,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;AAAA,MACzB,IAAA,EAAM,OAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AACA,IAAA,MAAM,qBAAA,GAAwB;AAAA,MAC5B,IAAA,EAAM,WAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AAEA,IAAA,KAAA,CAAM;AAAA,MACJ,WAAA,EAAa;AAAA,QACX,IAAA,EAAM,WAAA,EAAa,IAAA,IAAQ,kBAAA,CAAmB,IAAA;AAAA,QAC9C,MAAA,EAAQ,WAAA,EAAa,MAAA,IAAU,kBAAA,CAAmB;AAAA,OACpD;AAAA,MACA,cAAA,EAAgB;AAAA,QACd,IAAA,EAAM,cAAA,EAAgB,IAAA,IAAQ,qBAAA,CAAsB,IAAA;AAAA,QACpD,MAAA,EAAQ,cAAA,EAAgB,MAAA,IAAU,qBAAA,CAAsB;AAAA,OAC1D;AAAA,MACA,SAAS,OAAA,IAAW;AAAA,KACrB,CAAA;AAED,IAAA,MAAM,YAAA,GAAe,aAAa,MAAA,IAAU,aAAA;AAC5C,IAAA,IAAI,CAAC,YAAA,EAAc;AACjB,MAAA,MAAM,IAAI,MAAM,sCAAsC,CAAA;AAAA,IACxD;AACA,IAAA,IAAA,CAAK,eAAe,IAAIC,uBAAA,CAAO,EAAE,MAAA,EAAQ,cAAc,CAAA;AAEvD,IAAA,MAAM,eAAA,GAAkB,gBAAgB,MAAA,IAAU,aAAA;AAClD,IAAA,IAAI,CAAC,eAAA,EAAiB;AACpB,MAAA,MAAM,IAAI,MAAM,yCAAyC,CAAA;AAAA,IAC3D;AACA,IAAA,IAAA,CAAK,kBAAkB,IAAIA,uBAAA,CAAO,EAAE,MAAA,EAAQ,iBAAiB,CAAA;AAE7D,IAAA,IAAI,CAAC,IAAA,CAAK,YAAA,IAAgB,CAAC,KAAK,eAAA,EAAiB;AAC/C,MAAA,MAAM,IAAI,MAAM,0FAA0F,CAAA;AAAA,IAC5G;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EASA,MAAM,WAAA,GAA0D;AAC9D,IAAA,IAAI,CAAC,KAAK,WAAA,EAAa;AACrB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,OAAO;AAAA,MACL,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,SAAA,EAAU;AAAA,MACrB,EAAE,SAAS,KAAA,EAAM;AAAA,MACjB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA;AAAO,KACpB;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,KAAA,CACJ,KAAA,EACA,OAAA,EAKgC;AAChC,IAAA,IAAI,CAAC,KAAK,YAAA,EAAc;AACtB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,MAAA,MAAM,SAAmB,EAAC;AAC1B,MAAA,WAAA,MAAiB,SAAS,KAAA,EAAO;AAC/B,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,KAAA,GAAQ,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA,CAAE,SAAS,OAAO,CAAA;AAAA,IAChD;AAEA,IAAA,IAAI,KAAA,CAAM,IAAA,EAAK,CAAE,MAAA,KAAW,CAAA,EAAG;AAC7B,MAAA,MAAM,IAAI,MAAM,qBAAqB,CAAA;AAAA,IACvC;AAEA,IAAA,MAAM,EAAE,SAAS,cAAA,EAAgB,KAAA,EAAO,GAAG,YAAA,EAAa,GAAI,WAAW,EAAC;AAExE,IAAA,MAAM,KAAA,GAAQ,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AAC1C,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,YAAA,CAAc,KAAA,CAAM,OAAO,MAAA,CAAO;AAAA,QAC5D,KAAA,EAAO,IAAA,CAAK,WAAA,EAAa,IAAA,IAAQ,OAAA;AAAA,QACjC,KAAA,EAAQ,WAAW,IAAA,CAAK,OAAA;AAAA,QACxB,iBAAiB,cAAA,IAAkB,KAAA;AAAA,QACnC,KAAA;AAAA,QACA,OAAO,KAAA,IAAS,CAAA;AAAA,QAChB,GAAG;AAAA,OACJ,CAAA;AAED,MAAA,MAAM,WAAA,GAAc,IAAIC,kBAAA,EAAY;AACpC,MAAA,MAAM,SAAS,MAAA,CAAO,IAAA,CAAK,MAAM,QAAA,CAAS,aAAa,CAAA;AACvD,MAAA,WAAA,CAAY,IAAI,MAAM,CAAA;AACtB,MAAA,OAAO,WAAA;AAAA,IACT,CAAA,EAAG,oBAAoB,CAAA,EAAE;AAEzB,IAAA,OAAO,KAAA;AAAA,EACT;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAOA,MAAM,WAAA,GAAc;AAClB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,OAAO,EAAE,SAAS,KAAA,EAAM;AAAA,IAC1B;AACA,IAAA,OAAO,EAAE,SAAS,IAAA,EAAK;AAAA,EACzB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,MAAA,CACJ,WAAA,EACA,OAAA,EAIiB;AACjB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,MAAM,IAAI,MAAM,gCAAgC,CAAA;AAAA,IAClD;AAEA,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,WAAA,GAAc,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA;AAExC,IAAA,MAAM,IAAA,GAAO,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AACzC,MAAA,MAAM,EAAE,QAAA,EAAU,GAAG,YAAA,EAAa,GAAI,WAAW,EAAC;AAClD,MAAA,MAAM,IAAA,GAAO,IAAI,IAAA,CAAK,CAAC,WAAW,CAAA,EAAG,CAAA,MAAA,EAAS,QAAA,IAAY,KAAK,CAAA,CAAE,CAAA;AAEjE,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,eAAA,CAAiB,KAAA,CAAM,eAAe,MAAA,CAAO;AAAA,QACvE,KAAA,EAAO,IAAA,CAAK,cAAA,EAAgB,IAAA,IAAQ,WAAA;AAAA,QACpC,IAAA;AAAA,QACA,GAAG;AAAA,OACJ,CAAA;AAED,MAAA,OAAO,QAAA,CAAS,IAAA;AAAA,IAClB,CAAA,EAAG,qBAAqB,CAAA,EAAE;AAE1B,IAAA,OAAO,IAAA;AAAA,EACT;AACF","file":"index.cjs","sourcesContent":["import { PassThrough } from 'stream';\n\nimport { MastraVoice } from '@mastra/core/voice';\nimport OpenAI from 'openai';\n\ntype OpenAIVoiceId = 'alloy' | 'echo' | 'fable' | 'onyx' | 'nova' | 'shimmer' | 'ash' | 'coral' | 'sage';\ntype OpenAIModel = 'tts-1' | 'tts-1-hd' | 'whisper-1';\n\nexport interface OpenAIConfig {\n  name?: OpenAIModel;\n  apiKey?: string;\n}\n\nexport interface OpenAIVoiceConfig {\n  speech?: {\n    model: 'tts-1' | 'tts-1-hd';\n    apiKey?: string;\n    speaker?: OpenAIVoiceId;\n  };\n  listening?: {\n    model: 'whisper-1';\n    apiKey?: string;\n  };\n}\n\nexport class OpenAIVoice extends MastraVoice {\n  speechClient?: OpenAI;\n  listeningClient?: OpenAI;\n\n  /**\n   * Constructs an instance of OpenAIVoice with optional configurations for speech and listening models.\n   *\n   * @param {Object} [config] - Configuration options for the OpenAIVoice instance.\n   * @param {OpenAIConfig} [config.listeningModel] - Configuration for the listening model, including model name and API key.\n   * @param {OpenAIConfig} [config.speechModel] - Configuration for the speech model, including model name and API key.\n   * @param {string} [config.speaker] - The default speaker's voice to use for speech synthesis.\n   * @throws {Error} - Throws an error if no API key is provided for either the speech or listening model.\n   */\n  constructor({\n    listeningModel,\n    speechModel,\n    speaker,\n  }: {\n    listeningModel?: OpenAIConfig;\n    speechModel?: OpenAIConfig;\n    speaker?: string;\n  } = {}) {\n    const defaultApiKey = process.env.OPENAI_API_KEY;\n    const defaultSpeechModel = {\n      name: 'tts-1',\n      apiKey: defaultApiKey,\n    };\n    const defaultListeningModel = {\n      name: 'whisper-1',\n      apiKey: defaultApiKey,\n    };\n\n    super({\n      speechModel: {\n        name: speechModel?.name ?? defaultSpeechModel.name,\n        apiKey: speechModel?.apiKey ?? defaultSpeechModel.apiKey,\n      },\n      listeningModel: {\n        name: listeningModel?.name ?? defaultListeningModel.name,\n        apiKey: listeningModel?.apiKey ?? defaultListeningModel.apiKey,\n      },\n      speaker: speaker ?? 'alloy',\n    });\n\n    const speechApiKey = speechModel?.apiKey || defaultApiKey;\n    if (!speechApiKey) {\n      throw new Error('No API key provided for speech model');\n    }\n    this.speechClient = new OpenAI({ apiKey: speechApiKey });\n\n    const listeningApiKey = listeningModel?.apiKey || defaultApiKey;\n    if (!listeningApiKey) {\n      throw new Error('No API key provided for listening model');\n    }\n    this.listeningClient = new OpenAI({ apiKey: listeningApiKey });\n\n    if (!this.speechClient && !this.listeningClient) {\n      throw new Error('At least one of OPENAI_API_KEY, speechModel.apiKey, or listeningModel.apiKey must be set');\n    }\n  }\n\n  /**\n   * Retrieves a list of available speakers for the speech model.\n   *\n   * @returns {Promise<Array<{ voiceId: OpenAIVoiceId }>>} - A promise that resolves to an array of objects,\n   * each containing a `voiceId` representing an available speaker.\n   * @throws {Error} - Throws an error if the speech model is not configured.\n   */\n  async getSpeakers(): Promise<Array<{ voiceId: OpenAIVoiceId }>> {\n    if (!this.speechModel) {\n      throw new Error('Speech model not configured');\n    }\n\n    return [\n      { voiceId: 'alloy' },\n      { voiceId: 'echo' },\n      { voiceId: 'fable' },\n      { voiceId: 'onyx' },\n      { voiceId: 'nova' },\n      { voiceId: 'shimmer' },\n      { voiceId: 'ash' },\n      { voiceId: 'coral' },\n      { voiceId: 'sage' },\n    ];\n  }\n\n  /**\n   * Converts text or audio input into speech using the configured speech model.\n   *\n   * @param {string | NodeJS.ReadableStream} input - The text or audio stream to be converted into speech.\n   * @param {Object} [options] - Optional parameters for the speech synthesis.\n   * @param {string} [options.speaker] - The speaker's voice to use for the speech synthesis.\n   * @param {number} [options.speed] - The speed at which the speech should be synthesized.\n   * @returns {Promise<NodeJS.ReadableStream>} - A promise that resolves to a readable stream of the synthesized audio.\n   * @throws {Error} - Throws an error if the speech model is not configured or if the input text is empty.\n   */\n  async speak(\n    input: string | NodeJS.ReadableStream,\n    options?: {\n      speaker?: string;\n      speed?: number;\n      [key: string]: any;\n    },\n  ): Promise<NodeJS.ReadableStream> {\n    if (!this.speechClient) {\n      throw new Error('Speech model not configured');\n    }\n\n    if (typeof input !== 'string') {\n      const chunks: Buffer[] = [];\n      for await (const chunk of input) {\n        if (typeof chunk === 'string') {\n          chunks.push(Buffer.from(chunk));\n        } else {\n          chunks.push(chunk);\n        }\n      }\n      input = Buffer.concat(chunks).toString('utf-8');\n    }\n\n    if (input.trim().length === 0) {\n      throw new Error('Input text is empty');\n    }\n\n    const { speaker, responseFormat, speed, ...otherOptions } = options || {};\n\n    const audio = await this.traced(async () => {\n      const response = await this.speechClient!.audio.speech.create({\n        model: this.speechModel?.name ?? 'tts-1',\n        voice: (speaker ?? this.speaker) as OpenAIVoiceId,\n        response_format: responseFormat ?? 'mp3',\n        input,\n        speed: speed || 1.0,\n        ...otherOptions,\n      });\n\n      const passThrough = new PassThrough();\n      const buffer = Buffer.from(await response.arrayBuffer());\n      passThrough.end(buffer);\n      return passThrough;\n    }, 'voice.openai.speak')();\n\n    return audio;\n  }\n\n  /**\n   * Checks if listening capabilities are enabled.\n   *\n   * @returns {Promise<{ enabled: boolean }>}\n   */\n  async getListener() {\n    if (!this.listeningClient) {\n      return { enabled: false };\n    }\n    return { enabled: true };\n  }\n\n  /**\n   * Transcribes audio from a given stream using the configured listening model.\n   *\n   * @param {NodeJS.ReadableStream} audioStream - The audio stream to be transcribed.\n   * @param {Object} [options] - Optional parameters for the transcription.\n   * @param {string} [options.filetype] - The file type of the audio stream.\n   *                                      Supported types include 'mp3', 'mp4', 'mpeg', 'mpga', 'm4a', 'wav', 'webm'.\n   * @returns {Promise<string>} - A promise that resolves to the transcribed text.\n   * @throws {Error} - Throws an error if the listening model is not configured.\n   */\n  async listen(\n    audioStream: NodeJS.ReadableStream,\n    options?: {\n      filetype?: 'mp3' | 'mp4' | 'mpeg' | 'mpga' | 'm4a' | 'wav' | 'webm';\n      [key: string]: any;\n    },\n  ): Promise<string> {\n    if (!this.listeningClient) {\n      throw new Error('Listening model not configured');\n    }\n\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 audioBuffer = Buffer.concat(chunks);\n\n    const text = await this.traced(async () => {\n      const { filetype, ...otherOptions } = options || {};\n      const file = new File([audioBuffer], `audio.${filetype || 'mp3'}`);\n\n      const response = await this.listeningClient!.audio.transcriptions.create({\n        model: this.listeningModel?.name || 'whisper-1',\n        file: file as any,\n        ...otherOptions,\n      });\n\n      return response.text;\n    }, 'voice.openai.listen')();\n\n    return text;\n  }\n}\n"]}

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAEA,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,MAAM,MAAM,QAAQ,CAAC;AAE5B,KAAK,aAAa,GAAG,OAAO,GAAG,MAAM,GAAG,OAAO,GAAG,MAAM,GAAG,MAAM,GAAG,SAAS,GAAG,KAAK,GAAG,OAAO,GAAG,MAAM,CAAC;AACzG,KAAK,WAAW,GAAG,OAAO,GAAG,UAAU,GAAG,WAAW,CAAC;AAEtD,MAAM,WAAW,YAAY;IAC3B,IAAI,CAAC,EAAE,WAAW,CAAC;IACnB,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,iBAAiB;IAChC,MAAM,CAAC,EAAE;QACP,KAAK,EAAE,OAAO,GAAG,UAAU,CAAC;QAC5B,MAAM,CAAC,EAAE,MAAM,CAAC;QAChB,OAAO,CAAC,EAAE,aAAa,CAAC;KACzB,CAAC;IACF,SAAS,CAAC,EAAE;QACV,KAAK,EAAE,WAAW,CAAC;QACnB,MAAM,CAAC,EAAE,MAAM,CAAC;KACjB,CAAC;CACH;AAED,qBAAa,WAAY,SAAQ,WAAW;IAC1C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,eAAe,CAAC,EAAE,MAAM,CAAC;IAEzB;;;;;;;;OAQG;gBACS,EACV,cAAc,EACd,WAAW,EACX,OAAO,GACR,GAAE;QACD,cAAc,CAAC,EAAE,YAAY,CAAC;QAC9B,WAAW,CAAC,EAAE,YAAY,CAAC;QAC3B,OAAO,CAAC,EAAE,MAAM,CAAC;KACb;IAwCN;;;;;;OAMG;IACG,WAAW,IAAI,OAAO,CAAC,KAAK,CAAC;QAAE,OAAO,EAAE,aAAa,CAAA;KAAE,CAAC,CAAC;IAkB/D;;;;;;;;;OASG;IACG,KAAK,CACT,KAAK,EAAE,MAAM,GAAG,MAAM,CAAC,cAAc,EACrC,OAAO,CAAC,EAAE;QACR,OAAO,CAAC,EAAE,MAAM,CAAC;QACjB,KAAK,CAAC,EAAE,MAAM,CAAC;QACf,CAAC,GAAG,EAAE,MAAM,GAAG,GAAG,CAAC;KACpB,GACA,OAAO,CAAC,MAAM,CAAC,cAAc,CAAC;IAuCjC;;;;OAIG;IACG,WAAW;;;IAOjB;;;;;;;;;OASG;IACG,MAAM,CACV,WAAW,EAAE,MAAM,CAAC,cAAc,EAClC,OAAO,CAAC,EAAE;QACR,QAAQ,CAAC,EAAE,KAAK,GAAG,KAAK,GAAG,MAAM,GAAG,MAAM,GAAG,KAAK,GAAG,KAAK,GAAG,MAAM,CAAC;QACpE,CAAC,GAAG,EAAE,MAAM,GAAG,GAAG,CAAC;KACpB,GACA,OAAO,CAAC,MAAM,CAAC;CA8BnB"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAEA,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,MAAM,MAAM,QAAQ,CAAC;AAE5B,KAAK,aAAa,GAAG,OAAO,GAAG,MAAM,GAAG,OAAO,GAAG,MAAM,GAAG,MAAM,GAAG,SAAS,GAAG,KAAK,GAAG,OAAO,GAAG,MAAM,CAAC;AACzG,KAAK,WAAW,GAAG,OAAO,GAAG,UAAU,GAAG,WAAW,CAAC;AAEtD,MAAM,WAAW,YAAY;IAC3B,IAAI,CAAC,EAAE,WAAW,CAAC;IACnB,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,iBAAiB;IAChC,MAAM,CAAC,EAAE;QACP,KAAK,EAAE,OAAO,GAAG,UAAU,CAAC;QAC5B,MAAM,CAAC,EAAE,MAAM,CAAC;QAChB,OAAO,CAAC,EAAE,aAAa,CAAC;KACzB,CAAC;IACF,SAAS,CAAC,EAAE;QACV,KAAK,EAAE,WAAW,CAAC;QACnB,MAAM,CAAC,EAAE,MAAM,CAAC;KACjB,CAAC;CACH;AAED,qBAAa,WAAY,SAAQ,WAAW;IAC1C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,eAAe,CAAC,EAAE,MAAM,CAAC;IAEzB;;;;;;;;OAQG;gBACS,EACV,cAAc,EACd,WAAW,EACX,OAAO,GACR,GAAE;QACD,cAAc,CAAC,EAAE,YAAY,CAAC;QAC9B,WAAW,CAAC,EAAE,YAAY,CAAC;QAC3B,OAAO,CAAC,EAAE,MAAM,CAAC;KACb;IAwCN;;;;;;OAMG;IACG,WAAW,IAAI,OAAO,CAAC,KAAK,CAAC;QAAE,OAAO,EAAE,aAAa,CAAA;KAAE,CAAC,CAAC;IAkB/D;;;;;;;;;OASG;IACG,KAAK,CACT,KAAK,EAAE,MAAM,GAAG,MAAM,CAAC,cAAc,EACrC,OAAO,CAAC,EAAE;QACR,OAAO,CAAC,EAAE,MAAM,CAAC;QACjB,KAAK,CAAC,EAAE,MAAM,CAAC;QACf,CAAC,GAAG,EAAE,MAAM,GAAG,GAAG,CAAC;KACpB,GACA,OAAO,CAAC,MAAM,CAAC,cAAc,CAAC;IA0CjC;;;;OAIG;IACG,WAAW;;;IAOjB;;;;;;;;;OASG;IACG,MAAM,CACV,WAAW,EAAE,MAAM,CAAC,cAAc,EAClC,OAAO,CAAC,EAAE;QACR,QAAQ,CAAC,EAAE,KAAK,GAAG,KAAK,GAAG,MAAM,GAAG,MAAM,GAAG,KAAK,GAAG,KAAK,GAAG,MAAM,CAAC;QACpE,CAAC,GAAG,EAAE,MAAM,GAAG,GAAG,CAAC;KACpB,GACA,OAAO,CAAC,MAAM,CAAC;CA8BnB"}

package/dist/index.js

@@ -105,13 +105,15 @@ var OpenAIVoice = class extends MastraVoice {
     if (input.trim().length === 0) {
       throw new Error("Input text is empty");
     }
+    const { speaker, responseFormat, speed, ...otherOptions } = options || {};
     const audio = await this.traced(async () => {
       const response = await this.speechClient.audio.speech.create({
         model: this.speechModel?.name ?? "tts-1",
-        voice: options?.speaker ?? this.speaker,
-        response_format: options?.responseFormat ?? "mp3",
+        voice: speaker ?? this.speaker,
+        response_format: responseFormat ?? "mp3",
         input,
-        speed: options?.speed || 1
+        speed: speed || 1,
+        ...otherOptions
       });
       const passThrough = new PassThrough();
       const buffer = Buffer.from(await response.arrayBuffer());

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["../src/index.ts"],"names":[],"mappings":";;;;;AAyBO,IAAM,WAAA,GAAN,cAA0B,WAAA,CAAY;AAAA,EAC3C,YAAA;AAAA,EACA,eAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAWA,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;AAAA,MACzB,IAAA,EAAM,OAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AACA,IAAA,MAAM,qBAAA,GAAwB;AAAA,MAC5B,IAAA,EAAM,WAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AAEA,IAAA,KAAA,CAAM;AAAA,MACJ,WAAA,EAAa;AAAA,QACX,IAAA,EAAM,WAAA,EAAa,IAAA,IAAQ,kBAAA,CAAmB,IAAA;AAAA,QAC9C,MAAA,EAAQ,WAAA,EAAa,MAAA,IAAU,kBAAA,CAAmB;AAAA,OACpD;AAAA,MACA,cAAA,EAAgB;AAAA,QACd,IAAA,EAAM,cAAA,EAAgB,IAAA,IAAQ,qBAAA,CAAsB,IAAA;AAAA,QACpD,MAAA,EAAQ,cAAA,EAAgB,MAAA,IAAU,qBAAA,CAAsB;AAAA,OAC1D;AAAA,MACA,SAAS,OAAA,IAAW;AAAA,KACrB,CAAA;AAED,IAAA,MAAM,YAAA,GAAe,aAAa,MAAA,IAAU,aAAA;AAC5C,IAAA,IAAI,CAAC,YAAA,EAAc;AACjB,MAAA,MAAM,IAAI,MAAM,sCAAsC,CAAA;AAAA,IACxD;AACA,IAAA,IAAA,CAAK,eAAe,IAAI,MAAA,CAAO,EAAE,MAAA,EAAQ,cAAc,CAAA;AAEvD,IAAA,MAAM,eAAA,GAAkB,gBAAgB,MAAA,IAAU,aAAA;AAClD,IAAA,IAAI,CAAC,eAAA,EAAiB;AACpB,MAAA,MAAM,IAAI,MAAM,yCAAyC,CAAA;AAAA,IAC3D;AACA,IAAA,IAAA,CAAK,kBAAkB,IAAI,MAAA,CAAO,EAAE,MAAA,EAAQ,iBAAiB,CAAA;AAE7D,IAAA,IAAI,CAAC,IAAA,CAAK,YAAA,IAAgB,CAAC,KAAK,eAAA,EAAiB;AAC/C,MAAA,MAAM,IAAI,MAAM,0FAA0F,CAAA;AAAA,IAC5G;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EASA,MAAM,WAAA,GAA0D;AAC9D,IAAA,IAAI,CAAC,KAAK,WAAA,EAAa;AACrB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,OAAO;AAAA,MACL,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,SAAA,EAAU;AAAA,MACrB,EAAE,SAAS,KAAA,EAAM;AAAA,MACjB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA;AAAO,KACpB;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,KAAA,CACJ,KAAA,EACA,OAAA,EAKgC;AAChC,IAAA,IAAI,CAAC,KAAK,YAAA,EAAc;AACtB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,MAAA,MAAM,SAAmB,EAAC;AAC1B,MAAA,WAAA,MAAiB,SAAS,KAAA,EAAO;AAC/B,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,KAAA,GAAQ,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA,CAAE,SAAS,OAAO,CAAA;AAAA,IAChD;AAEA,IAAA,IAAI,KAAA,CAAM,IAAA,EAAK,CAAE,MAAA,KAAW,CAAA,EAAG;AAC7B,MAAA,MAAM,IAAI,MAAM,qBAAqB,CAAA;AAAA,IACvC;AAEA,IAAA,MAAM,KAAA,GAAQ,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AAC1C,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,YAAA,CAAc,KAAA,CAAM,OAAO,MAAA,CAAO;AAAA,QAC5D,KAAA,EAAO,IAAA,CAAK,WAAA,EAAa,IAAA,IAAQ,OAAA;AAAA,QACjC,KAAA,EAAQ,OAAA,EAAS,OAAA,IAAW,IAAA,CAAK,OAAA;AAAA,QACjC,eAAA,EAAiB,SAAS,cAAA,IAAkB,KAAA;AAAA,QAC5C,KAAA;AAAA,QACA,KAAA,EAAO,SAAS,KAAA,IAAS;AAAA,OAC1B,CAAA;AAED,MAAA,MAAM,WAAA,GAAc,IAAI,WAAA,EAAY;AACpC,MAAA,MAAM,SAAS,MAAA,CAAO,IAAA,CAAK,MAAM,QAAA,CAAS,aAAa,CAAA;AACvD,MAAA,WAAA,CAAY,IAAI,MAAM,CAAA;AACtB,MAAA,OAAO,WAAA;AAAA,IACT,CAAA,EAAG,oBAAoB,CAAA,EAAE;AAEzB,IAAA,OAAO,KAAA;AAAA,EACT;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAOA,MAAM,WAAA,GAAc;AAClB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,OAAO,EAAE,SAAS,KAAA,EAAM;AAAA,IAC1B;AACA,IAAA,OAAO,EAAE,SAAS,IAAA,EAAK;AAAA,EACzB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,MAAA,CACJ,WAAA,EACA,OAAA,EAIiB;AACjB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,MAAM,IAAI,MAAM,gCAAgC,CAAA;AAAA,IAClD;AAEA,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,WAAA,GAAc,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA;AAExC,IAAA,MAAM,IAAA,GAAO,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AACzC,MAAA,MAAM,EAAE,QAAA,EAAU,GAAG,YAAA,EAAa,GAAI,WAAW,EAAC;AAClD,MAAA,MAAM,IAAA,GAAO,IAAI,IAAA,CAAK,CAAC,WAAW,CAAA,EAAG,CAAA,MAAA,EAAS,QAAA,IAAY,KAAK,CAAA,CAAE,CAAA;AAEjE,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,eAAA,CAAiB,KAAA,CAAM,eAAe,MAAA,CAAO;AAAA,QACvE,KAAA,EAAO,IAAA,CAAK,cAAA,EAAgB,IAAA,IAAQ,WAAA;AAAA,QACpC,IAAA;AAAA,QACA,GAAG;AAAA,OACJ,CAAA;AAED,MAAA,OAAO,QAAA,CAAS,IAAA;AAAA,IAClB,CAAA,EAAG,qBAAqB,CAAA,EAAE;AAE1B,IAAA,OAAO,IAAA;AAAA,EACT;AACF","file":"index.js","sourcesContent":["import { PassThrough } from 'stream';\n\nimport { MastraVoice } from '@mastra/core/voice';\nimport OpenAI from 'openai';\n\ntype OpenAIVoiceId = 'alloy' | 'echo' | 'fable' | 'onyx' | 'nova' | 'shimmer' | 'ash' | 'coral' | 'sage';\ntype OpenAIModel = 'tts-1' | 'tts-1-hd' | 'whisper-1';\n\nexport interface OpenAIConfig {\n  name?: OpenAIModel;\n  apiKey?: string;\n}\n\nexport interface OpenAIVoiceConfig {\n  speech?: {\n    model: 'tts-1' | 'tts-1-hd';\n    apiKey?: string;\n    speaker?: OpenAIVoiceId;\n  };\n  listening?: {\n    model: 'whisper-1';\n    apiKey?: string;\n  };\n}\n\nexport class OpenAIVoice extends MastraVoice {\n  speechClient?: OpenAI;\n  listeningClient?: OpenAI;\n\n  /**\n   * Constructs an instance of OpenAIVoice with optional configurations for speech and listening models.\n   *\n   * @param {Object} [config] - Configuration options for the OpenAIVoice instance.\n   * @param {OpenAIConfig} [config.listeningModel] - Configuration for the listening model, including model name and API key.\n   * @param {OpenAIConfig} [config.speechModel] - Configuration for the speech model, including model name and API key.\n   * @param {string} [config.speaker] - The default speaker's voice to use for speech synthesis.\n   * @throws {Error} - Throws an error if no API key is provided for either the speech or listening model.\n   */\n  constructor({\n    listeningModel,\n    speechModel,\n    speaker,\n  }: {\n    listeningModel?: OpenAIConfig;\n    speechModel?: OpenAIConfig;\n    speaker?: string;\n  } = {}) {\n    const defaultApiKey = process.env.OPENAI_API_KEY;\n    const defaultSpeechModel = {\n      name: 'tts-1',\n      apiKey: defaultApiKey,\n    };\n    const defaultListeningModel = {\n      name: 'whisper-1',\n      apiKey: defaultApiKey,\n    };\n\n    super({\n      speechModel: {\n        name: speechModel?.name ?? defaultSpeechModel.name,\n        apiKey: speechModel?.apiKey ?? defaultSpeechModel.apiKey,\n      },\n      listeningModel: {\n        name: listeningModel?.name ?? defaultListeningModel.name,\n        apiKey: listeningModel?.apiKey ?? defaultListeningModel.apiKey,\n      },\n      speaker: speaker ?? 'alloy',\n    });\n\n    const speechApiKey = speechModel?.apiKey || defaultApiKey;\n    if (!speechApiKey) {\n      throw new Error('No API key provided for speech model');\n    }\n    this.speechClient = new OpenAI({ apiKey: speechApiKey });\n\n    const listeningApiKey = listeningModel?.apiKey || defaultApiKey;\n    if (!listeningApiKey) {\n      throw new Error('No API key provided for listening model');\n    }\n    this.listeningClient = new OpenAI({ apiKey: listeningApiKey });\n\n    if (!this.speechClient && !this.listeningClient) {\n      throw new Error('At least one of OPENAI_API_KEY, speechModel.apiKey, or listeningModel.apiKey must be set');\n    }\n  }\n\n  /**\n   * Retrieves a list of available speakers for the speech model.\n   *\n   * @returns {Promise<Array<{ voiceId: OpenAIVoiceId }>>} - A promise that resolves to an array of objects,\n   * each containing a `voiceId` representing an available speaker.\n   * @throws {Error} - Throws an error if the speech model is not configured.\n   */\n  async getSpeakers(): Promise<Array<{ voiceId: OpenAIVoiceId }>> {\n    if (!this.speechModel) {\n      throw new Error('Speech model not configured');\n    }\n\n    return [\n      { voiceId: 'alloy' },\n      { voiceId: 'echo' },\n      { voiceId: 'fable' },\n      { voiceId: 'onyx' },\n      { voiceId: 'nova' },\n      { voiceId: 'shimmer' },\n      { voiceId: 'ash' },\n      { voiceId: 'coral' },\n      { voiceId: 'sage' },\n    ];\n  }\n\n  /**\n   * Converts text or audio input into speech using the configured speech model.\n   *\n   * @param {string | NodeJS.ReadableStream} input - The text or audio stream to be converted into speech.\n   * @param {Object} [options] - Optional parameters for the speech synthesis.\n   * @param {string} [options.speaker] - The speaker's voice to use for the speech synthesis.\n   * @param {number} [options.speed] - The speed at which the speech should be synthesized.\n   * @returns {Promise<NodeJS.ReadableStream>} - A promise that resolves to a readable stream of the synthesized audio.\n   * @throws {Error} - Throws an error if the speech model is not configured or if the input text is empty.\n   */\n  async speak(\n    input: string | NodeJS.ReadableStream,\n    options?: {\n      speaker?: string;\n      speed?: number;\n      [key: string]: any;\n    },\n  ): Promise<NodeJS.ReadableStream> {\n    if (!this.speechClient) {\n      throw new Error('Speech model not configured');\n    }\n\n    if (typeof input !== 'string') {\n      const chunks: Buffer[] = [];\n      for await (const chunk of input) {\n        if (typeof chunk === 'string') {\n          chunks.push(Buffer.from(chunk));\n        } else {\n          chunks.push(chunk);\n        }\n      }\n      input = Buffer.concat(chunks).toString('utf-8');\n    }\n\n    if (input.trim().length === 0) {\n      throw new Error('Input text is empty');\n    }\n\n    const audio = await this.traced(async () => {\n      const response = await this.speechClient!.audio.speech.create({\n        model: this.speechModel?.name ?? 'tts-1',\n        voice: (options?.speaker ?? this.speaker) as OpenAIVoiceId,\n        response_format: options?.responseFormat ?? 'mp3',\n        input,\n        speed: options?.speed || 1.0,\n      });\n\n      const passThrough = new PassThrough();\n      const buffer = Buffer.from(await response.arrayBuffer());\n      passThrough.end(buffer);\n      return passThrough;\n    }, 'voice.openai.speak')();\n\n    return audio;\n  }\n\n  /**\n   * Checks if listening capabilities are enabled.\n   *\n   * @returns {Promise<{ enabled: boolean }>}\n   */\n  async getListener() {\n    if (!this.listeningClient) {\n      return { enabled: false };\n    }\n    return { enabled: true };\n  }\n\n  /**\n   * Transcribes audio from a given stream using the configured listening model.\n   *\n   * @param {NodeJS.ReadableStream} audioStream - The audio stream to be transcribed.\n   * @param {Object} [options] - Optional parameters for the transcription.\n   * @param {string} [options.filetype] - The file type of the audio stream.\n   *                                      Supported types include 'mp3', 'mp4', 'mpeg', 'mpga', 'm4a', 'wav', 'webm'.\n   * @returns {Promise<string>} - A promise that resolves to the transcribed text.\n   * @throws {Error} - Throws an error if the listening model is not configured.\n   */\n  async listen(\n    audioStream: NodeJS.ReadableStream,\n    options?: {\n      filetype?: 'mp3' | 'mp4' | 'mpeg' | 'mpga' | 'm4a' | 'wav' | 'webm';\n      [key: string]: any;\n    },\n  ): Promise<string> {\n    if (!this.listeningClient) {\n      throw new Error('Listening model not configured');\n    }\n\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 audioBuffer = Buffer.concat(chunks);\n\n    const text = await this.traced(async () => {\n      const { filetype, ...otherOptions } = options || {};\n      const file = new File([audioBuffer], `audio.${filetype || 'mp3'}`);\n\n      const response = await this.listeningClient!.audio.transcriptions.create({\n        model: this.listeningModel?.name || 'whisper-1',\n        file: file as any,\n        ...otherOptions,\n      });\n\n      return response.text;\n    }, 'voice.openai.listen')();\n\n    return text;\n  }\n}\n"]}
+{"version":3,"sources":["../src/index.ts"],"names":[],"mappings":";;;;;AAyBO,IAAM,WAAA,GAAN,cAA0B,WAAA,CAAY;AAAA,EAC3C,YAAA;AAAA,EACA,eAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAWA,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;AAAA,MACzB,IAAA,EAAM,OAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AACA,IAAA,MAAM,qBAAA,GAAwB;AAAA,MAC5B,IAAA,EAAM,WAAA;AAAA,MACN,MAAA,EAAQ;AAAA,KACV;AAEA,IAAA,KAAA,CAAM;AAAA,MACJ,WAAA,EAAa;AAAA,QACX,IAAA,EAAM,WAAA,EAAa,IAAA,IAAQ,kBAAA,CAAmB,IAAA;AAAA,QAC9C,MAAA,EAAQ,WAAA,EAAa,MAAA,IAAU,kBAAA,CAAmB;AAAA,OACpD;AAAA,MACA,cAAA,EAAgB;AAAA,QACd,IAAA,EAAM,cAAA,EAAgB,IAAA,IAAQ,qBAAA,CAAsB,IAAA;AAAA,QACpD,MAAA,EAAQ,cAAA,EAAgB,MAAA,IAAU,qBAAA,CAAsB;AAAA,OAC1D;AAAA,MACA,SAAS,OAAA,IAAW;AAAA,KACrB,CAAA;AAED,IAAA,MAAM,YAAA,GAAe,aAAa,MAAA,IAAU,aAAA;AAC5C,IAAA,IAAI,CAAC,YAAA,EAAc;AACjB,MAAA,MAAM,IAAI,MAAM,sCAAsC,CAAA;AAAA,IACxD;AACA,IAAA,IAAA,CAAK,eAAe,IAAI,MAAA,CAAO,EAAE,MAAA,EAAQ,cAAc,CAAA;AAEvD,IAAA,MAAM,eAAA,GAAkB,gBAAgB,MAAA,IAAU,aAAA;AAClD,IAAA,IAAI,CAAC,eAAA,EAAiB;AACpB,MAAA,MAAM,IAAI,MAAM,yCAAyC,CAAA;AAAA,IAC3D;AACA,IAAA,IAAA,CAAK,kBAAkB,IAAI,MAAA,CAAO,EAAE,MAAA,EAAQ,iBAAiB,CAAA;AAE7D,IAAA,IAAI,CAAC,IAAA,CAAK,YAAA,IAAgB,CAAC,KAAK,eAAA,EAAiB;AAC/C,MAAA,MAAM,IAAI,MAAM,0FAA0F,CAAA;AAAA,IAC5G;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EASA,MAAM,WAAA,GAA0D;AAC9D,IAAA,IAAI,CAAC,KAAK,WAAA,EAAa;AACrB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,OAAO;AAAA,MACL,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,MAAA,EAAO;AAAA,MAClB,EAAE,SAAS,SAAA,EAAU;AAAA,MACrB,EAAE,SAAS,KAAA,EAAM;AAAA,MACjB,EAAE,SAAS,OAAA,EAAQ;AAAA,MACnB,EAAE,SAAS,MAAA;AAAO,KACpB;AAAA,EACF;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,KAAA,CACJ,KAAA,EACA,OAAA,EAKgC;AAChC,IAAA,IAAI,CAAC,KAAK,YAAA,EAAc;AACtB,MAAA,MAAM,IAAI,MAAM,6BAA6B,CAAA;AAAA,IAC/C;AAEA,IAAA,IAAI,OAAO,UAAU,QAAA,EAAU;AAC7B,MAAA,MAAM,SAAmB,EAAC;AAC1B,MAAA,WAAA,MAAiB,SAAS,KAAA,EAAO;AAC/B,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,KAAA,GAAQ,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA,CAAE,SAAS,OAAO,CAAA;AAAA,IAChD;AAEA,IAAA,IAAI,KAAA,CAAM,IAAA,EAAK,CAAE,MAAA,KAAW,CAAA,EAAG;AAC7B,MAAA,MAAM,IAAI,MAAM,qBAAqB,CAAA;AAAA,IACvC;AAEA,IAAA,MAAM,EAAE,SAAS,cAAA,EAAgB,KAAA,EAAO,GAAG,YAAA,EAAa,GAAI,WAAW,EAAC;AAExE,IAAA,MAAM,KAAA,GAAQ,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AAC1C,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,YAAA,CAAc,KAAA,CAAM,OAAO,MAAA,CAAO;AAAA,QAC5D,KAAA,EAAO,IAAA,CAAK,WAAA,EAAa,IAAA,IAAQ,OAAA;AAAA,QACjC,KAAA,EAAQ,WAAW,IAAA,CAAK,OAAA;AAAA,QACxB,iBAAiB,cAAA,IAAkB,KAAA;AAAA,QACnC,KAAA;AAAA,QACA,OAAO,KAAA,IAAS,CAAA;AAAA,QAChB,GAAG;AAAA,OACJ,CAAA;AAED,MAAA,MAAM,WAAA,GAAc,IAAI,WAAA,EAAY;AACpC,MAAA,MAAM,SAAS,MAAA,CAAO,IAAA,CAAK,MAAM,QAAA,CAAS,aAAa,CAAA;AACvD,MAAA,WAAA,CAAY,IAAI,MAAM,CAAA;AACtB,MAAA,OAAO,WAAA;AAAA,IACT,CAAA,EAAG,oBAAoB,CAAA,EAAE;AAEzB,IAAA,OAAO,KAAA;AAAA,EACT;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAOA,MAAM,WAAA,GAAc;AAClB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,OAAO,EAAE,SAAS,KAAA,EAAM;AAAA,IAC1B;AACA,IAAA,OAAO,EAAE,SAAS,IAAA,EAAK;AAAA,EACzB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,EAYA,MAAM,MAAA,CACJ,WAAA,EACA,OAAA,EAIiB;AACjB,IAAA,IAAI,CAAC,KAAK,eAAA,EAAiB;AACzB,MAAA,MAAM,IAAI,MAAM,gCAAgC,CAAA;AAAA,IAClD;AAEA,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,WAAA,GAAc,MAAA,CAAO,MAAA,CAAO,MAAM,CAAA;AAExC,IAAA,MAAM,IAAA,GAAO,MAAM,IAAA,CAAK,MAAA,CAAO,YAAY;AACzC,MAAA,MAAM,EAAE,QAAA,EAAU,GAAG,YAAA,EAAa,GAAI,WAAW,EAAC;AAClD,MAAA,MAAM,IAAA,GAAO,IAAI,IAAA,CAAK,CAAC,WAAW,CAAA,EAAG,CAAA,MAAA,EAAS,QAAA,IAAY,KAAK,CAAA,CAAE,CAAA;AAEjE,MAAA,MAAM,WAAW,MAAM,IAAA,CAAK,eAAA,CAAiB,KAAA,CAAM,eAAe,MAAA,CAAO;AAAA,QACvE,KAAA,EAAO,IAAA,CAAK,cAAA,EAAgB,IAAA,IAAQ,WAAA;AAAA,QACpC,IAAA;AAAA,QACA,GAAG;AAAA,OACJ,CAAA;AAED,MAAA,OAAO,QAAA,CAAS,IAAA;AAAA,IAClB,CAAA,EAAG,qBAAqB,CAAA,EAAE;AAE1B,IAAA,OAAO,IAAA;AAAA,EACT;AACF","file":"index.js","sourcesContent":["import { PassThrough } from 'stream';\n\nimport { MastraVoice } from '@mastra/core/voice';\nimport OpenAI from 'openai';\n\ntype OpenAIVoiceId = 'alloy' | 'echo' | 'fable' | 'onyx' | 'nova' | 'shimmer' | 'ash' | 'coral' | 'sage';\ntype OpenAIModel = 'tts-1' | 'tts-1-hd' | 'whisper-1';\n\nexport interface OpenAIConfig {\n  name?: OpenAIModel;\n  apiKey?: string;\n}\n\nexport interface OpenAIVoiceConfig {\n  speech?: {\n    model: 'tts-1' | 'tts-1-hd';\n    apiKey?: string;\n    speaker?: OpenAIVoiceId;\n  };\n  listening?: {\n    model: 'whisper-1';\n    apiKey?: string;\n  };\n}\n\nexport class OpenAIVoice extends MastraVoice {\n  speechClient?: OpenAI;\n  listeningClient?: OpenAI;\n\n  /**\n   * Constructs an instance of OpenAIVoice with optional configurations for speech and listening models.\n   *\n   * @param {Object} [config] - Configuration options for the OpenAIVoice instance.\n   * @param {OpenAIConfig} [config.listeningModel] - Configuration for the listening model, including model name and API key.\n   * @param {OpenAIConfig} [config.speechModel] - Configuration for the speech model, including model name and API key.\n   * @param {string} [config.speaker] - The default speaker's voice to use for speech synthesis.\n   * @throws {Error} - Throws an error if no API key is provided for either the speech or listening model.\n   */\n  constructor({\n    listeningModel,\n    speechModel,\n    speaker,\n  }: {\n    listeningModel?: OpenAIConfig;\n    speechModel?: OpenAIConfig;\n    speaker?: string;\n  } = {}) {\n    const defaultApiKey = process.env.OPENAI_API_KEY;\n    const defaultSpeechModel = {\n      name: 'tts-1',\n      apiKey: defaultApiKey,\n    };\n    const defaultListeningModel = {\n      name: 'whisper-1',\n      apiKey: defaultApiKey,\n    };\n\n    super({\n      speechModel: {\n        name: speechModel?.name ?? defaultSpeechModel.name,\n        apiKey: speechModel?.apiKey ?? defaultSpeechModel.apiKey,\n      },\n      listeningModel: {\n        name: listeningModel?.name ?? defaultListeningModel.name,\n        apiKey: listeningModel?.apiKey ?? defaultListeningModel.apiKey,\n      },\n      speaker: speaker ?? 'alloy',\n    });\n\n    const speechApiKey = speechModel?.apiKey || defaultApiKey;\n    if (!speechApiKey) {\n      throw new Error('No API key provided for speech model');\n    }\n    this.speechClient = new OpenAI({ apiKey: speechApiKey });\n\n    const listeningApiKey = listeningModel?.apiKey || defaultApiKey;\n    if (!listeningApiKey) {\n      throw new Error('No API key provided for listening model');\n    }\n    this.listeningClient = new OpenAI({ apiKey: listeningApiKey });\n\n    if (!this.speechClient && !this.listeningClient) {\n      throw new Error('At least one of OPENAI_API_KEY, speechModel.apiKey, or listeningModel.apiKey must be set');\n    }\n  }\n\n  /**\n   * Retrieves a list of available speakers for the speech model.\n   *\n   * @returns {Promise<Array<{ voiceId: OpenAIVoiceId }>>} - A promise that resolves to an array of objects,\n   * each containing a `voiceId` representing an available speaker.\n   * @throws {Error} - Throws an error if the speech model is not configured.\n   */\n  async getSpeakers(): Promise<Array<{ voiceId: OpenAIVoiceId }>> {\n    if (!this.speechModel) {\n      throw new Error('Speech model not configured');\n    }\n\n    return [\n      { voiceId: 'alloy' },\n      { voiceId: 'echo' },\n      { voiceId: 'fable' },\n      { voiceId: 'onyx' },\n      { voiceId: 'nova' },\n      { voiceId: 'shimmer' },\n      { voiceId: 'ash' },\n      { voiceId: 'coral' },\n      { voiceId: 'sage' },\n    ];\n  }\n\n  /**\n   * Converts text or audio input into speech using the configured speech model.\n   *\n   * @param {string | NodeJS.ReadableStream} input - The text or audio stream to be converted into speech.\n   * @param {Object} [options] - Optional parameters for the speech synthesis.\n   * @param {string} [options.speaker] - The speaker's voice to use for the speech synthesis.\n   * @param {number} [options.speed] - The speed at which the speech should be synthesized.\n   * @returns {Promise<NodeJS.ReadableStream>} - A promise that resolves to a readable stream of the synthesized audio.\n   * @throws {Error} - Throws an error if the speech model is not configured or if the input text is empty.\n   */\n  async speak(\n    input: string | NodeJS.ReadableStream,\n    options?: {\n      speaker?: string;\n      speed?: number;\n      [key: string]: any;\n    },\n  ): Promise<NodeJS.ReadableStream> {\n    if (!this.speechClient) {\n      throw new Error('Speech model not configured');\n    }\n\n    if (typeof input !== 'string') {\n      const chunks: Buffer[] = [];\n      for await (const chunk of input) {\n        if (typeof chunk === 'string') {\n          chunks.push(Buffer.from(chunk));\n        } else {\n          chunks.push(chunk);\n        }\n      }\n      input = Buffer.concat(chunks).toString('utf-8');\n    }\n\n    if (input.trim().length === 0) {\n      throw new Error('Input text is empty');\n    }\n\n    const { speaker, responseFormat, speed, ...otherOptions } = options || {};\n\n    const audio = await this.traced(async () => {\n      const response = await this.speechClient!.audio.speech.create({\n        model: this.speechModel?.name ?? 'tts-1',\n        voice: (speaker ?? this.speaker) as OpenAIVoiceId,\n        response_format: responseFormat ?? 'mp3',\n        input,\n        speed: speed || 1.0,\n        ...otherOptions,\n      });\n\n      const passThrough = new PassThrough();\n      const buffer = Buffer.from(await response.arrayBuffer());\n      passThrough.end(buffer);\n      return passThrough;\n    }, 'voice.openai.speak')();\n\n    return audio;\n  }\n\n  /**\n   * Checks if listening capabilities are enabled.\n   *\n   * @returns {Promise<{ enabled: boolean }>}\n   */\n  async getListener() {\n    if (!this.listeningClient) {\n      return { enabled: false };\n    }\n    return { enabled: true };\n  }\n\n  /**\n   * Transcribes audio from a given stream using the configured listening model.\n   *\n   * @param {NodeJS.ReadableStream} audioStream - The audio stream to be transcribed.\n   * @param {Object} [options] - Optional parameters for the transcription.\n   * @param {string} [options.filetype] - The file type of the audio stream.\n   *                                      Supported types include 'mp3', 'mp4', 'mpeg', 'mpga', 'm4a', 'wav', 'webm'.\n   * @returns {Promise<string>} - A promise that resolves to the transcribed text.\n   * @throws {Error} - Throws an error if the listening model is not configured.\n   */\n  async listen(\n    audioStream: NodeJS.ReadableStream,\n    options?: {\n      filetype?: 'mp3' | 'mp4' | 'mpeg' | 'mpga' | 'm4a' | 'wav' | 'webm';\n      [key: string]: any;\n    },\n  ): Promise<string> {\n    if (!this.listeningClient) {\n      throw new Error('Listening model not configured');\n    }\n\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 audioBuffer = Buffer.concat(chunks);\n\n    const text = await this.traced(async () => {\n      const { filetype, ...otherOptions } = options || {};\n      const file = new File([audioBuffer], `audio.${filetype || 'mp3'}`);\n\n      const response = await this.listeningClient!.audio.transcriptions.create({\n        model: this.listeningModel?.name || 'whisper-1',\n        file: file as any,\n        ...otherOptions,\n      });\n\n      return response.text;\n    }, 'voice.openai.listen')();\n\n    return text;\n  }\n}\n"]}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mastra/voice-openai",
-  "version": "0.11.6",
+  "version": "0.11.7-alpha.0",
   "description": "Mastra OpenAI speech integration",
   "type": "module",
   "files": [
@@ -33,12 +33,12 @@
     "tsup": "^8.5.0",
     "typescript": "^5.8.3",
     "vitest": "^3.2.4",
+    "@mastra/core": "0.19.0-alpha.1",
     "@internal/lint": "0.0.42",
-    "@internal/types-builder": "0.0.17",
-    "@mastra/core": "0.18.0"
+    "@internal/types-builder": "0.0.17"
   },
   "peerDependencies": {
-    "@mastra/core": ">=0.15.3-0 <0.19.0-0",
+    "@mastra/core": ">=0.18.1-0 <0.20.0-0",
     "zod": "^3.25.0 || ^4.0.0"
   },
   "homepage": "https://mastra.ai",