@mastra/voice-openai 0.1.2-alpha.1 → 0.1.2-alpha.3

package/.turbo/turbo-build.log

@@ -1,18 +1,23 @@
 
-> @mastra/voice-openai@0.1.2-alpha.1 build /home/runner/work/mastra/mastra/voice/openai
-> tsup src/index.ts --format esm --experimental-dts --clean --treeshake
+> @mastra/voice-openai@0.1.2-alpha.3 build /home/runner/work/mastra/mastra/voice/openai
+> tsup src/index.ts --format esm,cjs --experimental-dts --clean --treeshake
 
 CLI Building entry: src/index.ts
 CLI Using tsconfig: tsconfig.json
 CLI tsup v8.3.6
 TSC Build start
-TSC ⚡️ Build success in 7687ms
+TSC ⚡️ Build success in 6267ms
 DTS Build start
 CLI Target: es2022
 Analysis will use the bundled TypeScript version 5.7.3
 Writing package typings: /home/runner/work/mastra/mastra/voice/openai/dist/_tsup-dts-rollup.d.ts
-DTS ⚡️ Build success in 6240ms
+Analysis will use the bundled TypeScript version 5.7.3
+Writing package typings: /home/runner/work/mastra/mastra/voice/openai/dist/_tsup-dts-rollup.d.cts
+DTS ⚡️ Build success in 10839ms
 CLI Cleaning output folder
 ESM Build start
+CJS Build start
 ESM dist/index.js 5.80 KB
-ESM ⚡️ Build success in 308ms
+ESM ⚡️ Build success in 540ms
+CJS dist/index.cjs 6.00 KB
+CJS ⚡️ Build success in 530ms

package/CHANGELOG.md

@@ -1,5 +1,26 @@
 # @mastra/voice-openai
 
+## 0.1.2-alpha.3
+
+### Patch Changes
+
+- bb4f447: Add support for commonjs
+- Updated dependencies [0fd78ac]
+- Updated dependencies [0d25b75]
+- Updated dependencies [fd14a3f]
+- Updated dependencies [3f369a2]
+- Updated dependencies [4d4e1e1]
+- Updated dependencies [bb4f447]
+  - @mastra/core@0.4.3-alpha.3
+
+## 0.1.2-alpha.2
+
+### Patch Changes
+
+- Updated dependencies [2512a93]
+- Updated dependencies [e62de74]
+  - @mastra/core@0.4.3-alpha.2
+
 ## 0.1.2-alpha.1
 
 ### Patch Changes

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

@@ -0,0 +1,83 @@
+import { MastraVoice } from '@mastra/core/voice';
+import OpenAI from 'openai';
+
+export declare interface OpenAIConfig {
+    name?: OpenAIModel;
+    apiKey?: string;
+}
+
+declare type OpenAIModel = 'tts-1' | 'tts-1-hd' | 'whisper-1';
+
+export declare class OpenAIVoice extends MastraVoice {
+    speechClient?: OpenAI;
+    listeningClient?: OpenAI;
+    /**
+     * Constructs an instance of OpenAIVoice with optional configurations for speech and listening models.
+     *
+     * @param {Object} [config] - Configuration options for the OpenAIVoice instance.
+     * @param {OpenAIConfig} [config.listeningModel] - Configuration for the listening model, including model name and API key.
+     * @param {OpenAIConfig} [config.speechModel] - Configuration for the speech model, including model name and API key.
+     * @param {string} [config.speaker] - The default speaker's voice to use for speech synthesis.
+     * @throws {Error} - Throws an error if no API key is provided for either the speech or listening model.
+     */
+    constructor({ listeningModel, speechModel, speaker, }?: {
+        listeningModel?: OpenAIConfig;
+        speechModel?: OpenAIConfig;
+        speaker?: string;
+    });
+    /**
+     * Retrieves a list of available speakers for the speech model.
+     *
+     * @returns {Promise<Array<{ voiceId: OpenAIVoiceId }>>} - A promise that resolves to an array of objects,
+     * each containing a `voiceId` representing an available speaker.
+     * @throws {Error} - Throws an error if the speech model is not configured.
+     */
+    getSpeakers(): Promise<Array<{
+        voiceId: OpenAIVoiceId;
+    }>>;
+    /**
+     * Converts text or audio input into speech using the configured speech model.
+     *
+     * @param {string | NodeJS.ReadableStream} input - The text or audio stream to be converted into speech.
+     * @param {Object} [options] - Optional parameters for the speech synthesis.
+     * @param {string} [options.speaker] - The speaker's voice to use for the speech synthesis.
+     * @param {number} [options.speed] - The speed at which the speech should be synthesized.
+     * @returns {Promise<NodeJS.ReadableStream>} - A promise that resolves to a readable stream of the synthesized audio.
+     * @throws {Error} - Throws an error if the speech model is not configured or if the input text is empty.
+     */
+    speak(input: string | NodeJS.ReadableStream, options?: {
+        speaker?: string;
+        speed?: number;
+        [key: string]: any;
+    }): Promise<NodeJS.ReadableStream>;
+    /**
+     * Transcribes audio from a given stream using the configured listening model.
+     *
+     * @param {NodeJS.ReadableStream} audioStream - The audio stream to be transcribed.
+     * @param {Object} [options] - Optional parameters for the transcription.
+     * @param {string} [options.filetype] - The file type of the audio stream.
+     *                                      Supported types include 'mp3', 'mp4', 'mpeg', 'mpga', 'm4a', 'wav', 'webm'.
+     * @returns {Promise<string>} - A promise that resolves to the transcribed text.
+     * @throws {Error} - Throws an error if the listening model is not configured.
+     */
+    listen(audioStream: NodeJS.ReadableStream, options?: {
+        filetype?: 'mp3' | 'mp4' | 'mpeg' | 'mpga' | 'm4a' | 'wav' | 'webm';
+        [key: string]: any;
+    }): Promise<string>;
+}
+
+export declare interface OpenAIVoiceConfig {
+    speech?: {
+        model: 'tts-1' | 'tts-1-hd';
+        apiKey?: string;
+        speaker?: OpenAIVoiceId;
+    };
+    listening?: {
+        model: 'whisper-1';
+        apiKey?: string;
+    };
+}
+
+declare type OpenAIVoiceId = 'alloy' | 'echo' | 'fable' | 'onyx' | 'nova' | 'shimmer' | 'ash' | 'coral' | 'sage';
+
+export { }

package/dist/index.cjs

@@ -0,0 +1,157 @@
+'use strict';
+
+var stream = require('stream');
+var voice = require('@mastra/core/voice');
+var OpenAI = require('openai');
+
+function _interopDefault (e) { return e && e.__esModule ? e : { default: e }; }
+
+var OpenAI__default = /*#__PURE__*/_interopDefault(OpenAI);
+
+// src/index.ts
+var OpenAIVoice = class extends voice.MastraVoice {
+  speechClient;
+  listeningClient;
+  /**
+   * Constructs an instance of OpenAIVoice with optional configurations for speech and listening models.
+   *
+   * @param {Object} [config] - Configuration options for the OpenAIVoice instance.
+   * @param {OpenAIConfig} [config.listeningModel] - Configuration for the listening model, including model name and API key.
+   * @param {OpenAIConfig} [config.speechModel] - Configuration for the speech model, including model name and API key.
+   * @param {string} [config.speaker] - The default speaker's voice to use for speech synthesis.
+   * @throws {Error} - Throws an error if no API key is provided for either the speech or listening model.
+   */
+  constructor({
+    listeningModel,
+    speechModel,
+    speaker
+  } = {}) {
+    const defaultApiKey = process.env.OPENAI_API_KEY;
+    const defaultSpeechModel = {
+      name: "tts-1",
+      apiKey: defaultApiKey
+    };
+    const defaultListeningModel = {
+      name: "whisper-1",
+      apiKey: defaultApiKey
+    };
+    super({
+      speechModel: {
+        name: speechModel?.name ?? defaultSpeechModel.name,
+        apiKey: speechModel?.apiKey ?? defaultSpeechModel.apiKey
+      },
+      listeningModel: {
+        name: listeningModel?.name ?? defaultListeningModel.name,
+        apiKey: listeningModel?.apiKey ?? defaultListeningModel.apiKey
+      },
+      speaker: speaker ?? "alloy"
+    });
+    const speechApiKey = speechModel?.apiKey || defaultApiKey;
+    if (!speechApiKey) {
+      throw new Error("No API key provided for speech model");
+    }
+    this.speechClient = new OpenAI__default.default({ apiKey: speechApiKey });
+    const listeningApiKey = listeningModel?.apiKey || defaultApiKey;
+    if (!listeningApiKey) {
+      throw new Error("No API key provided for listening model");
+    }
+    this.listeningClient = new OpenAI__default.default({ apiKey: listeningApiKey });
+    if (!this.speechClient && !this.listeningClient) {
+      throw new Error("At least one of OPENAI_API_KEY, speechModel.apiKey, or listeningModel.apiKey must be set");
+    }
+  }
+  /**
+   * Retrieves a list of available speakers for the speech model.
+   *
+   * @returns {Promise<Array<{ voiceId: OpenAIVoiceId }>>} - A promise that resolves to an array of objects,
+   * each containing a `voiceId` representing an available speaker.
+   * @throws {Error} - Throws an error if the speech model is not configured.
+   */
+  async getSpeakers() {
+    if (!this.speechModel) {
+      throw new Error("Speech model not configured");
+    }
+    return [
+      { voiceId: "alloy" },
+      { voiceId: "echo" },
+      { voiceId: "fable" },
+      { voiceId: "onyx" },
+      { voiceId: "nova" },
+      { voiceId: "shimmer" },
+      { voiceId: "ash" },
+      { voiceId: "coral" },
+      { voiceId: "sage" }
+    ];
+  }
+  /**
+   * Converts text or audio input into speech using the configured speech model.
+   *
+   * @param {string | NodeJS.ReadableStream} input - The text or audio stream to be converted into speech.
+   * @param {Object} [options] - Optional parameters for the speech synthesis.
+   * @param {string} [options.speaker] - The speaker's voice to use for the speech synthesis.
+   * @param {number} [options.speed] - The speed at which the speech should be synthesized.
+   * @returns {Promise<NodeJS.ReadableStream>} - A promise that resolves to a readable stream of the synthesized audio.
+   * @throws {Error} - Throws an error if the speech model is not configured or if the input text is empty.
+   */
+  async speak(input, options) {
+    if (!this.speechClient) {
+      throw new Error("Speech model not configured");
+    }
+    if (typeof input !== "string") {
+      const chunks = [];
+      for await (const chunk of input) {
+        chunks.push(Buffer.from(chunk));
+      }
+      input = Buffer.concat(chunks).toString("utf-8");
+    }
+    if (input.trim().length === 0) {
+      throw new Error("Input text is empty");
+    }
+    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,
+        input,
+        speed: options?.speed || 1
+      });
+      const passThrough = new stream.PassThrough();
+      const buffer = Buffer.from(await response.arrayBuffer());
+      passThrough.end(buffer);
+      return passThrough;
+    }, "voice.openai.speak")();
+    return audio;
+  }
+  /**
+   * Transcribes audio from a given stream using the configured listening model.
+   *
+   * @param {NodeJS.ReadableStream} audioStream - The audio stream to be transcribed.
+   * @param {Object} [options] - Optional parameters for the transcription.
+   * @param {string} [options.filetype] - The file type of the audio stream.
+   *                                      Supported types include 'mp3', 'mp4', 'mpeg', 'mpga', 'm4a', 'wav', 'webm'.
+   * @returns {Promise<string>} - A promise that resolves to the transcribed text.
+   * @throws {Error} - Throws an error if the listening model is not configured.
+   */
+  async listen(audioStream, options) {
+    if (!this.listeningClient) {
+      throw new Error("Listening model not configured");
+    }
+    const chunks = [];
+    for await (const chunk of audioStream) {
+      chunks.push(Buffer.from(chunk));
+    }
+    const audioBuffer = Buffer.concat(chunks);
+    const text = await this.traced(async () => {
+      const { filetype, ...otherOptions } = options || {};
+      const file = new File([audioBuffer], `audio.${filetype || "mp3"}`);
+      const response = await this.listeningClient.audio.transcriptions.create({
+        model: this.listeningModel?.name || "whisper-1",
+        file,
+        ...otherOptions
+      });
+      return response.text;
+    }, "voice.openai.listen")();
+    return text;
+  }
+};
+
+exports.OpenAIVoice = OpenAIVoice;

package/dist/index.d.cts

@@ -0,0 +1,3 @@
+export { OpenAIConfig } from './_tsup-dts-rollup.cjs';
+export { OpenAIVoiceConfig } from './_tsup-dts-rollup.cjs';
+export { OpenAIVoice } from './_tsup-dts-rollup.cjs';

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mastra/voice-openai",
-  "version": "0.1.2-alpha.1",
+  "version": "0.1.2-alpha.3",
   "description": "Mastra OpenAI speech integration",
   "type": "module",
   "main": "dist/index.js",
@@ -10,6 +10,10 @@
       "import": {
         "types": "./dist/index.d.ts",
         "default": "./dist/index.js"
+      },
+      "require": {
+        "types": "./dist/index.d.cts",
+        "default": "./dist/index.cjs"
       }
     },
     "./package.json": "./package.json"
@@ -17,7 +21,7 @@
   "dependencies": {
     "openai": "^4.28.0",
     "zod": "^3.24.1",
-    "@mastra/core": "^0.4.3-alpha.1"
+    "@mastra/core": "^0.4.3-alpha.3"
   },
   "devDependencies": {
     "@microsoft/api-extractor": "^7.49.2",
@@ -29,7 +33,7 @@
     "@internal/lint": "0.0.0"
   },
   "scripts": {
-    "build": "tsup src/index.ts --format esm --experimental-dts --clean --treeshake",
+    "build": "tsup src/index.ts --format esm,cjs --experimental-dts --clean --treeshake",
     "build:watch": "pnpm build --watch",
     "test": "vitest run",
     "lint": "eslint ."