phonic 0.5.0 → 0.6.0

package/README.md

@@ -7,7 +7,6 @@ Node.js library for the Phonic API.
 - [Usage](#usage)
   - [Get voices](#get-voices)
   - [Get voice by id](#get-voice-by-id)
-  - [Text-to-speech via WebSocket](#text-to-speech-via-websocket)
   - [Speech-to-speech via WebSocket](#speech-to-speech-via-websocket)
 
 ## Installation
@@ -49,7 +48,7 @@ if (error === null) {
 }
 ```
 
-### Speesh-to-speech via WebSocket
+### Speech-to-speech via WebSocket
 
 Open a WebSocket connection:
 
@@ -98,7 +97,7 @@ phonicWebSocket.onMessage((message) => {
 
     case "audio_chunk": {
       // Send the audio chunk to Twilio, for example:
-      ws.send(
+      twilioWebSocket.send(
         JSON.stringify({
           event: "media",
           streamSid: "...",
@@ -133,99 +132,6 @@ phonicWebSocket.onError((event) => {
 });
 ```
 
-### Text-to-speech via WebSocket
-
-Open a WebSocket connection:
-
-```ts
-const { data, error } = await phonic.tts.websocket({
-  model: "shasta",
-  output_format: "mulaw_8000",
-  voice_id: "meredith",
-});
-
-if (error !== null) {
-  throw new Error(error.message);
-}
-
-// Here we know that the WebSocket connection is open.
-const { phonicWebSocket } = data;
-```
-
-Process audio chunks that Phonic sends back to you, by sending them to Twilio, for example:
-
-```ts
-phonicWebSocket.onMessage((message) => {
-  if (message.type === "audio_chunk") {
-    ws.send(
-      JSON.stringify({
-        event: "media",
-        streamSid: "...",
-        media: {
-          payload: message.audio,
-        },
-      }),
-    );
-  }
-});
-```
-
-Send text chunks to Phonic for audio generation as you receive them from LLM:
-
-```ts
-const stream = await openai.chat.completions.create(...);
-
-for await (const chunk of stream) {
-  const text = chunk.choices[0]?.delta?.content || "";
-
-  if (text) {
-    phonicWebSocket.generate({ text });
-  }
-}
-```
-
-Tell Phonic to finish generating audio for all text chunks you've sent:
-
-```ts
-phonicWebSocket.flush();
-```
-
-You can also tell Phonic to stop sending audio chunks back, e.g. if the user interrupts the conversation:
-
-```ts
-phonicWebSocket.stop();
-```
-
-To close the WebSocket connection:
-
-```ts
-phonicWebSocket.close();
-```
-
-To know when the last audio chunk has been received:
-
-```ts
-phonicWebSocket.onMessage((message) => {
-  if (message.type === "flushed") {
-    console.log("Last audio chunk received");
-  }
-});
-```
-
-You can also listen for close and error events:
-
-```ts
-phonicWebSocket.onClose((event) => {
-  console.log(
-    `Phonic WebSocket closed with code ${event.code} and reason "${event.reason}"`,
-  );
-});
-
-phonicWebSocket.onError((event) => {
-  console.log(`Error from Phonic WebSocket: ${event.message}`);
-});
-```
-
 ## Publish a new version on npm
 
 1. `bunx changeset`

package/dist/index.d.mts

@@ -18,24 +18,13 @@ type DataOrError<T> = Promise<{
     error: ErrorResponse;
 }>;
 
-type PhonicTTSWebSocketParams = {
-    model?: string;
-    output_format?: string;
-    voice_id?: string;
-};
-type PhonicTTSWebSocketResponseMessage = {
-    type: "config";
-    model: string;
-    output_format: string;
-    voice_id: string;
+type PhonicSTSWebSocketResponseMessage = {
+    type: "input_text";
+    text: string;
 } | {
     type: "audio_chunk";
-    audio: string;
     text: string;
-} | {
-    type: "flush_confirm";
-} | {
-    type: "stop_confirm";
+    audio: string;
 } | {
     type: "error";
     error: {
@@ -43,40 +32,44 @@ type PhonicTTSWebSocketResponseMessage = {
         code?: string;
     };
     paramErrors?: {
-        model?: string;
-        output_format?: string;
+        system_prompt?: string;
+        welcome_message?: string;
         voice_id?: string;
-        text?: string;
-        speed?: string;
+        input_format?: string;
+        output_format?: string;
     };
 };
-type OnMessageCallback$1 = (message: PhonicTTSWebSocketResponseMessage) => void;
-type OnCloseCallback$1 = (event: WebSocket.CloseEvent) => void;
-type OnErrorCallback$1 = (event: WebSocket.ErrorEvent) => void;
+type OnMessageCallback = (message: PhonicSTSWebSocketResponseMessage) => void;
+type OnCloseCallback = (event: WebSocket.CloseEvent) => void;
+type OnErrorCallback = (event: WebSocket.ErrorEvent) => void;
 
-declare class PhonicTTSWebSocket {
+declare class PhonicSTSWebSocket {
     private readonly ws;
     private onMessageCallback;
     private onCloseCallback;
     private onErrorCallback;
     constructor(ws: WebSocket);
-    onMessage(callback: OnMessageCallback$1): void;
-    onClose(callback: OnCloseCallback$1): void;
-    onError(callback: OnErrorCallback$1): void;
-    generate(message: {
-        text: string;
-        speed?: number;
+    onMessage(callback: OnMessageCallback): void;
+    onClose(callback: OnCloseCallback): void;
+    onError(callback: OnErrorCallback): void;
+    config(message: {
+        system_prompt?: string;
+        welcome_message?: string;
+        voice_id?: string;
+        input_format?: "pcm_44100" | "mulaw_8000";
+        output_format?: "pcm_44100" | "mulaw_8000";
+    }): void;
+    audioChunk(message: {
+        audio: string;
     }): void;
-    flush(): void;
-    stop(): void;
     close(): void;
 }
 
-declare class TextToSpeech {
+declare class SpeechToSpeech {
     private readonly phonic;
     constructor(phonic: Phonic);
-    websocket(params?: PhonicTTSWebSocketParams): DataOrError<{
-        phonicWebSocket: PhonicTTSWebSocket;
+    websocket(): DataOrError<{
+        phonicWebSocket: PhonicSTSWebSocket;
     }>;
 }
 
@@ -105,7 +98,7 @@ declare class Phonic {
     readonly baseUrl: string;
     private readonly headers;
     readonly voices: Voices;
-    readonly tts: TextToSpeech;
+    readonly sts: SpeechToSpeech;
     constructor(apiKey: string, config?: PhonicConfig);
     fetchRequest<T>(path: string, options: FetchOptions): DataOrError<T>;
     get<T>(path: string): Promise<{
@@ -117,51 +110,4 @@ declare class Phonic {
     }>;
 }
 
-type PhonicSTSWebSocketResponseMessage = {
-    type: "input_text";
-    text: string;
-} | {
-    type: "audio_chunk";
-    text: string;
-    audio: string;
-} | {
-    type: "error";
-    error: {
-        message: string;
-        code?: string;
-    };
-    paramErrors?: {
-        system_prompt?: string;
-        welcome_message?: string;
-        voice_id?: string;
-        input_format?: string;
-        output_format?: string;
-    };
-};
-type OnMessageCallback = (message: PhonicSTSWebSocketResponseMessage) => void;
-type OnCloseCallback = (event: WebSocket.CloseEvent) => void;
-type OnErrorCallback = (event: WebSocket.ErrorEvent) => void;
-
-declare class PhonicSTSWebSocket {
-    private readonly ws;
-    private onMessageCallback;
-    private onCloseCallback;
-    private onErrorCallback;
-    constructor(ws: WebSocket);
-    onMessage(callback: OnMessageCallback): void;
-    onClose(callback: OnCloseCallback): void;
-    onError(callback: OnErrorCallback): void;
-    config(message: {
-        system_prompt?: string;
-        welcome_message?: string;
-        voice_id?: string;
-        input_format?: "pcm_44100" | "mulaw_8000";
-        output_format?: "pcm_44100" | "mulaw_8000";
-    }): void;
-    audioChunk(message: {
-        audio: string;
-    }): void;
-    close(): void;
-}
-
-export { Phonic, PhonicSTSWebSocket, PhonicTTSWebSocket };
+export { Phonic, PhonicSTSWebSocket };

package/dist/index.d.ts

@@ -18,24 +18,13 @@ type DataOrError<T> = Promise<{
     error: ErrorResponse;
 }>;
 
-type PhonicTTSWebSocketParams = {
-    model?: string;
-    output_format?: string;
-    voice_id?: string;
-};
-type PhonicTTSWebSocketResponseMessage = {
-    type: "config";
-    model: string;
-    output_format: string;
-    voice_id: string;
+type PhonicSTSWebSocketResponseMessage = {
+    type: "input_text";
+    text: string;
 } | {
     type: "audio_chunk";
-    audio: string;
     text: string;
-} | {
-    type: "flush_confirm";
-} | {
-    type: "stop_confirm";
+    audio: string;
 } | {
     type: "error";
     error: {
@@ -43,40 +32,44 @@ type PhonicTTSWebSocketResponseMessage = {
         code?: string;
     };
     paramErrors?: {
-        model?: string;
-        output_format?: string;
+        system_prompt?: string;
+        welcome_message?: string;
         voice_id?: string;
-        text?: string;
-        speed?: string;
+        input_format?: string;
+        output_format?: string;
     };
 };
-type OnMessageCallback$1 = (message: PhonicTTSWebSocketResponseMessage) => void;
-type OnCloseCallback$1 = (event: WebSocket.CloseEvent) => void;
-type OnErrorCallback$1 = (event: WebSocket.ErrorEvent) => void;
+type OnMessageCallback = (message: PhonicSTSWebSocketResponseMessage) => void;
+type OnCloseCallback = (event: WebSocket.CloseEvent) => void;
+type OnErrorCallback = (event: WebSocket.ErrorEvent) => void;
 
-declare class PhonicTTSWebSocket {
+declare class PhonicSTSWebSocket {
     private readonly ws;
     private onMessageCallback;
     private onCloseCallback;
     private onErrorCallback;
     constructor(ws: WebSocket);
-    onMessage(callback: OnMessageCallback$1): void;
-    onClose(callback: OnCloseCallback$1): void;
-    onError(callback: OnErrorCallback$1): void;
-    generate(message: {
-        text: string;
-        speed?: number;
+    onMessage(callback: OnMessageCallback): void;
+    onClose(callback: OnCloseCallback): void;
+    onError(callback: OnErrorCallback): void;
+    config(message: {
+        system_prompt?: string;
+        welcome_message?: string;
+        voice_id?: string;
+        input_format?: "pcm_44100" | "mulaw_8000";
+        output_format?: "pcm_44100" | "mulaw_8000";
+    }): void;
+    audioChunk(message: {
+        audio: string;
     }): void;
-    flush(): void;
-    stop(): void;
     close(): void;
 }
 
-declare class TextToSpeech {
+declare class SpeechToSpeech {
     private readonly phonic;
     constructor(phonic: Phonic);
-    websocket(params?: PhonicTTSWebSocketParams): DataOrError<{
-        phonicWebSocket: PhonicTTSWebSocket;
+    websocket(): DataOrError<{
+        phonicWebSocket: PhonicSTSWebSocket;
     }>;
 }
 
@@ -105,7 +98,7 @@ declare class Phonic {
     readonly baseUrl: string;
     private readonly headers;
     readonly voices: Voices;
-    readonly tts: TextToSpeech;
+    readonly sts: SpeechToSpeech;
     constructor(apiKey: string, config?: PhonicConfig);
     fetchRequest<T>(path: string, options: FetchOptions): DataOrError<T>;
     get<T>(path: string): Promise<{
@@ -117,51 +110,4 @@ declare class Phonic {
     }>;
 }
 
-type PhonicSTSWebSocketResponseMessage = {
-    type: "input_text";
-    text: string;
-} | {
-    type: "audio_chunk";
-    text: string;
-    audio: string;
-} | {
-    type: "error";
-    error: {
-        message: string;
-        code?: string;
-    };
-    paramErrors?: {
-        system_prompt?: string;
-        welcome_message?: string;
-        voice_id?: string;
-        input_format?: string;
-        output_format?: string;
-    };
-};
-type OnMessageCallback = (message: PhonicSTSWebSocketResponseMessage) => void;
-type OnCloseCallback = (event: WebSocket.CloseEvent) => void;
-type OnErrorCallback = (event: WebSocket.ErrorEvent) => void;
-
-declare class PhonicSTSWebSocket {
-    private readonly ws;
-    private onMessageCallback;
-    private onCloseCallback;
-    private onErrorCallback;
-    constructor(ws: WebSocket);
-    onMessage(callback: OnMessageCallback): void;
-    onClose(callback: OnCloseCallback): void;
-    onError(callback: OnErrorCallback): void;
-    config(message: {
-        system_prompt?: string;
-        welcome_message?: string;
-        voice_id?: string;
-        input_format?: "pcm_44100" | "mulaw_8000";
-        output_format?: "pcm_44100" | "mulaw_8000";
-    }): void;
-    audioChunk(message: {
-        audio: string;
-    }): void;
-    close(): void;
-}
-
-export { Phonic, PhonicSTSWebSocket, PhonicTTSWebSocket };
+export { Phonic, PhonicSTSWebSocket };

package/dist/index.js

@@ -35,13 +35,13 @@ __export(index_exports, {
 module.exports = __toCommonJS(index_exports);
 
 // package.json
-var version = "0.5.0";
+var version = "0.6.0";
 
-// src/tts/index.ts
+// src/sts/index.ts
 var import_ws = __toESM(require("ws"));
 
-// src/tts/websocket.ts
-var PhonicTTSWebSocket = class {
+// src/sts/websocket.ts
+var PhonicSTSWebSocket = class {
   constructor(ws) {
     this.ws = ws;
     this.ws.onmessage = (event) => {
@@ -71,9 +71,8 @@ var PhonicTTSWebSocket = class {
     this.onMessage = this.onMessage.bind(this);
     this.onClose = this.onClose.bind(this);
     this.onError = this.onError.bind(this);
-    this.generate = this.generate.bind(this);
-    this.flush = this.flush.bind(this);
-    this.stop = this.stop.bind(this);
+    this.config = this.config.bind(this);
+    this.audioChunk = this.audioChunk.bind(this);
     this.close = this.close.bind(this);
   }
   onMessageCallback = null;
@@ -88,41 +87,42 @@ var PhonicTTSWebSocket = class {
   onError(callback) {
     this.onErrorCallback = callback;
   }
-  generate(message) {
+  config(message) {
     this.ws.send(
       JSON.stringify({
-        type: "generate",
+        type: "config",
         ...message
       })
     );
   }
-  flush() {
-    this.ws.send(JSON.stringify({ type: "flush" }));
-  }
-  stop() {
-    this.ws.send(JSON.stringify({ type: "stop" }));
+  audioChunk(message) {
+    this.ws.send(
+      JSON.stringify({
+        type: "audio_chunk",
+        ...message
+      })
+    );
   }
   close() {
     this.ws.close();
   }
 };
 
-// src/tts/index.ts
-var TextToSpeech = class {
+// src/sts/index.ts
+var SpeechToSpeech = class {
   constructor(phonic) {
     this.phonic = phonic;
   }
-  async websocket(params) {
+  async websocket() {
     return new Promise((resolve) => {
       const wsBaseUrl = this.phonic.baseUrl.replace(/^http/, "ws");
-      const queryString = new URLSearchParams(params).toString();
-      const ws = new import_ws.default(`${wsBaseUrl}/v1/tts/ws?${queryString}`, {
+      const ws = new import_ws.default(`${wsBaseUrl}/v1/sts/ws`, {
         headers: {
           Authorization: `Bearer ${this.phonic.apiKey}`
         }
       });
       ws.onopen = () => {
-        const phonicWebSocket = new PhonicTTSWebSocket(ws);
+        const phonicWebSocket = new PhonicSTSWebSocket(ws);
         resolve({ data: { phonicWebSocket }, error: null });
       };
       ws.onerror = (error) => {
@@ -182,7 +182,7 @@ var Phonic = class {
   baseUrl;
   headers;
   voices = new Voices(this);
-  tts = new TextToSpeech(this);
+  sts = new SpeechToSpeech(this);
   async fetchRequest(path, options) {
     try {
       const response = await fetch(`${this.baseUrl}/v1${path}`, {

package/dist/index.mjs

@@ -1,11 +1,11 @@
 // package.json
-var version = "0.5.0";
+var version = "0.6.0";
 
-// src/tts/index.ts
+// src/sts/index.ts
 import WebSocket from "ws";
 
-// src/tts/websocket.ts
-var PhonicTTSWebSocket = class {
+// src/sts/websocket.ts
+var PhonicSTSWebSocket = class {
   constructor(ws) {
     this.ws = ws;
     this.ws.onmessage = (event) => {
@@ -35,9 +35,8 @@ var PhonicTTSWebSocket = class {
     this.onMessage = this.onMessage.bind(this);
     this.onClose = this.onClose.bind(this);
     this.onError = this.onError.bind(this);
-    this.generate = this.generate.bind(this);
-    this.flush = this.flush.bind(this);
-    this.stop = this.stop.bind(this);
+    this.config = this.config.bind(this);
+    this.audioChunk = this.audioChunk.bind(this);
     this.close = this.close.bind(this);
   }
   onMessageCallback = null;
@@ -52,41 +51,42 @@ var PhonicTTSWebSocket = class {
   onError(callback) {
     this.onErrorCallback = callback;
   }
-  generate(message) {
+  config(message) {
     this.ws.send(
       JSON.stringify({
-        type: "generate",
+        type: "config",
         ...message
       })
     );
   }
-  flush() {
-    this.ws.send(JSON.stringify({ type: "flush" }));
-  }
-  stop() {
-    this.ws.send(JSON.stringify({ type: "stop" }));
+  audioChunk(message) {
+    this.ws.send(
+      JSON.stringify({
+        type: "audio_chunk",
+        ...message
+      })
+    );
   }
   close() {
     this.ws.close();
   }
 };
 
-// src/tts/index.ts
-var TextToSpeech = class {
+// src/sts/index.ts
+var SpeechToSpeech = class {
   constructor(phonic) {
     this.phonic = phonic;
   }
-  async websocket(params) {
+  async websocket() {
     return new Promise((resolve) => {
       const wsBaseUrl = this.phonic.baseUrl.replace(/^http/, "ws");
-      const queryString = new URLSearchParams(params).toString();
-      const ws = new WebSocket(`${wsBaseUrl}/v1/tts/ws?${queryString}`, {
+      const ws = new WebSocket(`${wsBaseUrl}/v1/sts/ws`, {
         headers: {
           Authorization: `Bearer ${this.phonic.apiKey}`
         }
       });
       ws.onopen = () => {
-        const phonicWebSocket = new PhonicTTSWebSocket(ws);
+        const phonicWebSocket = new PhonicSTSWebSocket(ws);
         resolve({ data: { phonicWebSocket }, error: null });
       };
       ws.onerror = (error) => {
@@ -146,7 +146,7 @@ var Phonic = class {
   baseUrl;
   headers;
   voices = new Voices(this);
-  tts = new TextToSpeech(this);
+  sts = new SpeechToSpeech(this);
   async fetchRequest(path, options) {
     try {
       const response = await fetch(`${this.baseUrl}/v1${path}`, {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "phonic",
-  "version": "0.5.0",
+  "version": "0.6.0",
   "description": "Phonic Node.js SDK",
   "scripts": {
     "build": "tsup",
@@ -51,8 +51,7 @@
   },
   "keywords": [
     "phonic",
-    "text-to-speech",
-    "tts",
+    "speech-to-speech",
     "javascript",
     "typescript",
     "ai",