@cartesia/cartesia-js 1.0.0-alpha.4 → 1.0.1

package/.turbo/turbo-build.log

@@ -5,71 +5,71 @@ $ tsup src/ --format cjs,esm --dts
 CLI Target: es6
 CJS Build start
 ESM Build start
-ESM dist/index.js          437.00 B
+CJS dist/index.cjs         24.12 KB
+CJS dist/lib/client.cjs    3.34 KB
+CJS dist/react/index.cjs   31.68 KB
+CJS dist/tts/index.cjs     18.49 KB
+CJS dist/types/index.cjs   764.00 B
+CJS dist/voices/index.cjs  5.04 KB
+CJS dist/lib/constants.cjs 1.43 KB
+CJS dist/lib/index.cjs     20.02 KB
+CJS dist/react/utils.cjs   1.80 KB
+CJS dist/tts/player.cjs    6.96 KB
+CJS dist/tts/source.cjs    8.15 KB
+CJS dist/tts/utils.cjs     4.33 KB
+CJS dist/tts/websocket.cjs 18.18 KB
+CJS ⚡️ Build success in 110ms
 ESM dist/chunk-FXPGR372.js 0 B
+ESM dist/tts/source.js     144.00 B
+ESM dist/tts/utils.js      426.00 B
+ESM dist/tts/websocket.js  242.00 B
+ESM dist/types/index.js    31.00 B
+ESM dist/voices/index.js   174.00 B
+ESM dist/index.js          499.00 B
 ESM dist/lib/client.js     132.00 B
-ESM dist/react/index.js    6.22 KB
-ESM dist/tts/index.js      261.00 B
 ESM dist/lib/constants.js  183.00 B
 ESM dist/lib/index.js      322.00 B
-ESM dist/chunk-ISRU7PLL.js 353.00 B
-ESM dist/chunk-SGXUEFII.js 1.35 KB
+ESM dist/react/index.js    7.62 KB
+ESM dist/chunk-OFH3ML4L.js 353.00 B
+ESM dist/chunk-KUSVZXDT.js 1.35 KB
 ESM dist/react/utils.js    109.00 B
-ESM dist/chunk-3FL2SNIR.js 337.00 B
-ESM dist/chunk-VK7LBMVI.js 439.00 B
-ESM dist/chunk-IQAXBRHU.js 5.84 KB
-ESM dist/chunk-PQ6CIPFW.js 4.02 KB
-ESM dist/chunk-PQ5EVEEH.js 841.00 B
+ESM dist/chunk-NQVZNVOU.js 337.00 B
+ESM dist/tts/index.js      261.00 B
+ESM dist/chunk-2NA5SEML.js 439.00 B
+ESM dist/chunk-ASZKHN7Q.js 6.89 KB
+ESM dist/chunk-5M33ZF3Y.js 841.00 B
 ESM dist/chunk-2BFEKY3F.js 366.00 B
-ESM dist/tts/player.js     143.00 B
-ESM dist/chunk-36JBKJUN.js 3.52 KB
-ESM dist/chunk-RO7TY474.js 1.95 KB
-ESM dist/chunk-WIFMLPT5.js 2.27 KB
-ESM dist/types/index.js    31.00 B
-ESM dist/voices/index.js   174.00 B
-ESM dist/tts/source.js     112.00 B
-ESM dist/tts/utils.js      395.00 B
-ESM dist/tts/websocket.js  242.00 B
-ESM ⚡️ Build success in 100ms
-CJS dist/index.cjs         21.01 KB
-CJS dist/lib/client.cjs    3.34 KB
-CJS dist/lib/constants.cjs 1.43 KB
-CJS dist/lib/index.cjs     17.18 KB
-CJS dist/react/index.cjs   27.24 KB
-CJS dist/react/utils.cjs   1.80 KB
-CJS dist/tts/index.cjs     15.66 KB
-CJS dist/tts/player.cjs    6.66 KB
-CJS dist/tts/source.cjs    6.63 KB
-CJS dist/tts/utils.cjs     3.87 KB
-CJS dist/tts/websocket.cjs 15.34 KB
-CJS dist/types/index.cjs   764.00 B
-CJS dist/voices/index.cjs  5.04 KB
-CJS ⚡️ Build success in 105ms
+ESM dist/tts/player.js     174.00 B
+ESM dist/chunk-LZO6K34D.js 3.76 KB
+ESM dist/chunk-BHY7MNGT.js 2.11 KB
+ESM dist/chunk-6YQ6KDIQ.js 5.32 KB
+ESM dist/chunk-GHY2WEOK.js 2.70 KB
+ESM ⚡️ Build success in 112ms
 DTS Build start
-DTS ⚡️ Build success in 8445ms
-DTS dist/index.d.cts         509.00 B
+DTS ⚡️ Build success in 8520ms
+DTS dist/index.d.cts         804.00 B
 DTS dist/lib/constants.d.cts 564.00 B
 DTS dist/lib/index.d.cts     410.00 B
-DTS dist/react/index.d.cts   1018.00 B
+DTS dist/react/index.d.cts   1.07 KB
 DTS dist/react/utils.d.cts   240.00 B
 DTS dist/tts/index.d.cts     471.00 B
-DTS dist/tts/player.d.cts    1.06 KB
-DTS dist/tts/utils.d.cts     2.56 KB
-DTS dist/tts/websocket.d.cts 2.39 KB
-DTS dist/tts/source.d.cts    2.17 KB
+DTS dist/tts/player.d.cts    1.20 KB
+DTS dist/tts/utils.d.cts     2.68 KB
+DTS dist/tts/websocket.d.cts 2.94 KB
+DTS dist/tts/source.d.cts    2.55 KB
 DTS dist/voices/index.d.cts  399.00 B
 DTS dist/lib/client.d.cts    267.00 B
-DTS dist/types/index.d.cts   1.28 KB
-DTS dist/index.d.ts          501.00 B
+DTS dist/types/index.d.cts   3.07 KB
+DTS dist/index.d.ts          796.00 B
 DTS dist/lib/constants.d.ts  564.00 B
 DTS dist/lib/index.d.ts      404.00 B
-DTS dist/react/index.d.ts    1016.00 B
+DTS dist/react/index.d.ts    1.07 KB
 DTS dist/react/utils.d.ts    240.00 B
 DTS dist/tts/index.d.ts      467.00 B
-DTS dist/tts/player.d.ts     1.06 KB
-DTS dist/tts/utils.d.ts      2.55 KB
-DTS dist/tts/websocket.d.ts  2.39 KB
-DTS dist/tts/source.d.ts     2.16 KB
+DTS dist/tts/player.d.ts     1.20 KB
+DTS dist/tts/utils.d.ts      2.68 KB
+DTS dist/tts/websocket.d.ts  2.94 KB
+DTS dist/tts/source.d.ts     2.55 KB
 DTS dist/voices/index.d.ts   397.00 B
 DTS dist/lib/client.d.ts     266.00 B
-DTS dist/types/index.d.ts    1.28 KB
+DTS dist/types/index.d.ts    3.07 KB

package/CHANGELOG.md

@@ -1,5 +1,28 @@
 # @cartesia/cartesia-js
 
+## 1.0.1
+
+### Patch Changes
+
+- cb7adc2: Introduces support for continuations, timestamps, voice control and multiple output formats. Improves typing and error handling for the package.
+
+## 1.0.0
+
+### Major Changes
+
+- 3ee5bfc: Initial release of Cartesia client with voices and WebSocket support
+
+### Minor Changes
+
+- e49f73a: Stabilize audio playback in the browser to support play/pause functionality.
+
+### Patch Changes
+
+- c98a0c7: Fix typo in README
+- 38af01f: Fix how URLs are constructed, solving WebSocket connection failure
+- 8ecf940: Add provisional Node.js support
+- 585d2c9: Makes JS client compatible with the Cartesia Stable API (2024-06-10)
+
 ## 1.0.0-alpha.4
 
 ### Patch Changes

package/LICENSE.md

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Cartesia AI, Inc.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -1,4 +1,24 @@
-# JavaScript Client for Cartesia
+# Cartesia JavaScript Client
+
+![NPM Version](https://img.shields.io/npm/v/%40cartesia%2Fcartesia-js?logo=npm)
+[![Discord](https://badgen.net/badge/black/Cartesia/icon?icon=discord&label)](https://discord.gg/ZVxavqHB9X)
+
+This client provides convenient access to [Cartesia's TTS models](https://cartesia.ai/). Sonic is the fastest text-to-speech model around—it can generate a second of audio in just 650ms, and it can stream out the first audio chunk in just 135ms. Alongside Sonic, we also offer an extensive prebuilt voice library for a variety of use cases.
+
+The JavaScript client is a thin wrapper around the Cartesia API. You can view docs for the API at [docs.cartesia.ai](https://docs.cartesia.ai/).
+
+- [Cartesia JavaScript Client](#cartesia-javascript-client)
+	- [Installation](#installation)
+	- [Usage](#usage)
+		- [CRUD on Voices](#crud-on-voices)
+		- [TTS over WebSocket](#tts-over-websocket)
+			- [Input Streaming with Contexts](#input-streaming-with-contexts)
+			- [Timestamps](#timestamps)
+			- [Speed and emotion controls \[Alpha\]](#speed-and-emotion-controls-alpha)
+		- [Multilingual TTS \[Alpha\]](#multilingual-tts-alpha)
+		- [Playing audio in the browser](#playing-audio-in-the-browser)
+	- [React](#react)
+
 
 ## Installation
 
@@ -40,12 +60,6 @@ const newVoice = await cartesia.voices.create({
 });
 console.log(newVoice);
 
-// Clone a voice from a URL.
-const clonedVoice = await cartesia.voices.clone({
-	mode: "url",
-	url: "https://youtu.be/AdtLxlttrHg?si=07OLmDPg__0IN14f&t=6",
-});
-
 // Clone a voice from a file.
 const clonedVoice = await cartesia.voices.clone({
 	mode: "clip",
@@ -62,8 +76,12 @@ const cartesia = new Cartesia({
 	apiKey: "your-api-key",
 });
 
-// Initialize the WebSocket. Make sure the sample rate you specify is supported.
-const websocket = cartesia.tts.websocket({ sampleRate: 44100 });
+// Initialize the WebSocket. Make sure the output format you specify is supported.
+const websocket = cartesia.tts.websocket({
+	container: "raw",
+	encoding: "pcm_f32le",
+	sampleRate: 44100
+});
 
 try {
 	await websocket.connect();
@@ -73,14 +91,13 @@ try {
 
 // Create a stream.
 const response = await websocket.send({
-	model: "upbeat-moon",
+	model_id: "sonic-english",
 	voice: {
-		mode: "embedding",
-		embedding: Array(192).fill(1.0),
+		mode: "id",
+		id: "a0e99841-438c-4a64-b679-ae501e7d6091",
 	},
 	transcript: "Hello, world!"
 	// The WebSocket sets output_format on your behalf.
-	// The container is "raw" and the encoding is "pcm_f32le".
 });
 
 // Access the raw messages from the WebSocket.
@@ -96,12 +113,76 @@ for await (const message of response.events('message')) {
 }
 ```
 
-#### Playing audio in the browser
+#### Input Streaming with Contexts
+
+You can perform input streaming with contexts as described in the [docs](https://docs.cartesia.ai/api-reference/endpoints/stream-speech-websocket#input-streaming-with-contexts). The WebSocket's `send` method is just a wrapper around sending a message on the WebSocket, so the request format specified the docs can be used directly.
+
+You should use the return from the first `send` call on a context to receive outputs and events for the entire context. You can ignore the return values of subsequent `send` calls.
+
+#### Timestamps
+
+To receive timestamps in responses, set the `add_timestamps` field in the request object to `true`.
+
+```js
+const response = await websocket.send({
+	model_id: "sonic-english",
+	voice: {
+		mode: "id",
+		id: "a0e99841-438c-4a64-b679-ae501e7d6091",
+	},
+	transcript: "Hello, world!",
+	add_timestamps: true,
+});
+```
+
+You can then listen for timestamps on the returned response object.
+
+```js
+response.on("timestamps", (timestamps) => {
+	console.log("Received timestamps for words:", timestamps.words);
+	console.log("Words start at:", timestamps.start);
+	console.log("Words end at:", timestamps.end);
+});
+
+// You can also access timestamps using a for-await-of loop.
+for (await const timestamps of response.events('timestamps')) {
+	console.log("Received timestamps for words:", timestamps.words);
+	console.log("Words start at:", timestamps.start);
+	console.log("Words end at:", timestamps.end);
+}
+```
+
+#### Speed and emotion controls [Alpha]
+
+The API has experimental support for speed and emotion controls that is not subject to semantic versioning and is subject to change without notice. You can control the speed and emotion of the synthesized speech by setting the `speed` and `emotion` fields under `voice.__experimental_controls` in the request object.
+
+```js
+const response = await websocket.send({
+	model_id: "sonic-english",
+	voice: {
+		mode: "id",
+		id: "a0e99841-438c-4a64-b679-ae501e7d6091",
+		__experimental_controls: {
+			speed: "fastest",
+			emotion: ["sadness", "surprise:high"],
+		},
+	},
+	transcript: "Hello, world!",
+});
+```
+
+### Multilingual TTS [Alpha]
+
+You can define the language of the text you want to synthesize by setting the `language` field in the request object. Make sure that you are using `model_id: "sonic-multilingual"` in the request object.
+
+Supported languages are listed at [docs.cartesia.ai](https://docs.cartesia.ai/getting-started/available-models).
+
+### Playing audio in the browser
 
-(We currently only support playing audio in the browser. Support for other JS environments is coming soon.)
+(The `WebPlayer` class only supports playing audio in the browser and the raw PCM format with fp32le encoding.)
 
 ```js
-// If you're using the client in the browser, you can play the audio like this:
+// If you're using the client in the browser, you can control audio playback using our WebPlayer:
 import { WebPlayer } from "@cartesia/cartesia-js";
 
 console.log("Playing stream...");
@@ -118,7 +199,7 @@ console.log("Done playing.");
 
 ## React
 
-We export a React hook that simplifies the process of using the TTS API. The hook manages the WebSocket connection and provides a simple interface for buffering and playing audio.
+We export a React hook that simplifies the process of using the TTS API. The hook manages the WebSocket connection and provides a simple interface for buffering, playing, pausing and restarting audio.
 
 ```jsx
 import { useTTS } from '@cartesia/cartesia-js/react';
@@ -134,11 +215,11 @@ function TextToSpeech() {
 	const handlePlay = async () => {
 		// Begin buffering the audio.
 		const response = await tts.buffer({
-			model_id: "upbeat-moon",
+			model_id: "sonic-english",
 			voice: {
-				mode: "embedding",
-				embedding: Array(192).fill(1.0),
-			},
+        		mode: "id",
+        		id: "a0e99841-438c-4a64-b679-ae501e7d6091",
+        	},
 			transcript: text,
 		});
 

package/dist/{chunk-VK7LBMVI.js → chunk-2NA5SEML.js}

@@ -1,9 +1,9 @@
 import {
   WebSocket
-} from "./chunk-IQAXBRHU.js";
+} from "./chunk-ASZKHN7Q.js";
 import {
   Client
-} from "./chunk-PQ5EVEEH.js";
+} from "./chunk-5M33ZF3Y.js";
 
 // src/tts/index.ts
 var TTS = class extends Client {

package/dist/{chunk-PQ5EVEEH.js → chunk-5M33ZF3Y.js}

@@ -6,7 +6,7 @@ import {
 import {
   __spreadProps,
   __spreadValues
-} from "./chunk-WIFMLPT5.js";
+} from "./chunk-GHY2WEOK.js";
 
 // src/lib/client.ts
 import fetch from "cross-fetch";

package/dist/{chunk-PQ6CIPFW.js → chunk-6YQ6KDIQ.js}

@@ -2,12 +2,19 @@ import {
   __async,
   __privateAdd,
   __privateGet,
+  __privateMethod,
   __privateSet
-} from "./chunk-WIFMLPT5.js";
+} from "./chunk-GHY2WEOK.js";
 
 // src/tts/source.ts
 import Emittery from "emittery";
-var _emitter, _buffer, _readIndex, _writeIndex, _closed, _sampleRate;
+var ENCODING_MAP = {
+  pcm_f32le: { arrayType: Float32Array, bytesPerElement: 4 },
+  pcm_s16le: { arrayType: Int16Array, bytesPerElement: 2 },
+  pcm_alaw: { arrayType: Uint8Array, bytesPerElement: 1 },
+  pcm_mulaw: { arrayType: Uint8Array, bytesPerElement: 1 }
+};
+var _emitter, _buffer, _readIndex, _writeIndex, _closed, _sampleRate, _encoding, _container, _createBuffer, createBuffer_fn;
 var Source = class {
   /**
    * Create a new Source.
@@ -15,23 +22,44 @@ var Source = class {
    * @param options - Options for the Source.
    * @param options.sampleRate - The sample rate of the audio.
    */
-  constructor({ sampleRate }) {
+  constructor({
+    sampleRate,
+    encoding,
+    container
+  }) {
+    /**
+     * Create a new buffer for the source.
+     *
+     * @param size - The size of the buffer to create.
+     * @returns The new buffer as a TypedArray based on the encoding.
+     */
+    __privateAdd(this, _createBuffer);
     __privateAdd(this, _emitter, new Emittery());
     __privateAdd(this, _buffer, void 0);
     __privateAdd(this, _readIndex, 0);
     __privateAdd(this, _writeIndex, 0);
     __privateAdd(this, _closed, false);
     __privateAdd(this, _sampleRate, void 0);
+    __privateAdd(this, _encoding, void 0);
+    __privateAdd(this, _container, void 0);
     this.on = __privateGet(this, _emitter).on.bind(__privateGet(this, _emitter));
     this.once = __privateGet(this, _emitter).once.bind(__privateGet(this, _emitter));
     this.events = __privateGet(this, _emitter).events.bind(__privateGet(this, _emitter));
     this.off = __privateGet(this, _emitter).off.bind(__privateGet(this, _emitter));
     __privateSet(this, _sampleRate, sampleRate);
-    __privateSet(this, _buffer, new Float32Array(1024));
+    __privateSet(this, _encoding, encoding);
+    __privateSet(this, _container, container);
+    __privateSet(this, _buffer, __privateMethod(this, _createBuffer, createBuffer_fn).call(this, 1024));
   }
   get sampleRate() {
     return __privateGet(this, _sampleRate);
   }
+  get encoding() {
+    return __privateGet(this, _encoding);
+  }
+  get container() {
+    return __privateGet(this, _container);
+  }
   /**
    * Append audio to the buffer.
    *
@@ -45,7 +73,7 @@ var Source = class {
         while (newCapacity < requiredCapacity) {
           newCapacity *= 2;
         }
-        const newBuffer = new Float32Array(newCapacity);
+        const newBuffer = __privateMethod(this, _createBuffer, createBuffer_fn).call(this, newCapacity);
         newBuffer.set(__privateGet(this, _buffer));
         __privateSet(this, _buffer, newBuffer);
       }
@@ -93,6 +121,9 @@ var Source = class {
   get readIndex() {
     return __privateGet(this, _readIndex);
   }
+  get writeIndex() {
+    return __privateGet(this, _writeIndex);
+  }
   /**
    * Close the source. This signals that no more audio will be enqueued.
    *
@@ -114,7 +145,15 @@ _readIndex = new WeakMap();
 _writeIndex = new WeakMap();
 _closed = new WeakMap();
 _sampleRate = new WeakMap();
+_encoding = new WeakMap();
+_container = new WeakMap();
+_createBuffer = new WeakSet();
+createBuffer_fn = function(size) {
+  const { arrayType: ArrayType } = ENCODING_MAP[__privateGet(this, _encoding)];
+  return new ArrayType(size);
+};
 
 export {
+  ENCODING_MAP,
   Source
 };

package/dist/{chunk-IQAXBRHU.js → chunk-ASZKHN7Q.js}

@@ -1,9 +1,6 @@
-import {
-  Source
-} from "./chunk-PQ6CIPFW.js";
 import {
   Client
-} from "./chunk-PQ5EVEEH.js";
+} from "./chunk-5M33ZF3Y.js";
 import {
   CARTESIA_VERSION,
   constructApiUrl
@@ -13,29 +10,33 @@ import {
   createMessageHandlerForContextId,
   getEmitteryCallbacks,
   isSentinel
-} from "./chunk-RO7TY474.js";
+} from "./chunk-BHY7MNGT.js";
+import {
+  Source
+} from "./chunk-6YQ6KDIQ.js";
 import {
   __async,
+  __objRest,
   __privateAdd,
   __privateGet,
   __privateMethod,
   __privateSet,
   __spreadProps,
   __spreadValues
-} from "./chunk-WIFMLPT5.js";
+} from "./chunk-GHY2WEOK.js";
 
 // src/tts/websocket.ts
 import Emittery from "emittery";
 import { humanId } from "human-id";
 import { WebSocket as PartySocketWebSocket } from "partysocket";
-var _isConnected, _sampleRate, _generateId, generateId_fn;
+var _isConnected, _sampleRate, _container, _encoding, _generateId, generateId_fn;
 var WebSocket = class extends Client {
   /**
    * Create a new WebSocket client.
    *
    * @param args - Arguments to pass to the Client constructor.
    */
-  constructor({ sampleRate }, ...args) {
+  constructor({ sampleRate, container, encoding }, ...args) {
     super(...args);
     /**
      * Generate a unique ID suitable for a streaming context.
@@ -48,37 +49,47 @@ var WebSocket = class extends Client {
     __privateAdd(this, _generateId);
     __privateAdd(this, _isConnected, false);
     __privateAdd(this, _sampleRate, void 0);
+    __privateAdd(this, _container, void 0);
+    __privateAdd(this, _encoding, void 0);
     __privateSet(this, _sampleRate, sampleRate);
+    __privateSet(this, _container, container != null ? container : "raw");
+    __privateSet(this, _encoding, encoding != null ? encoding : "pcm_f32le");
   }
   /**
-   * Send a message over the WebSocket in order to start a stream.
+   * Send a message over the WebSocket to start a stream.
    *
-   * @param inputs - Stream options.
+   * @param inputs - Stream options. Defined in the StreamRequest type.
    * @param options - Options for the stream.
    * @param options.timeout - The maximum time to wait for a chunk before cancelling the stream.
-   * If `0`, the stream will not time out.
+   *                          If set to `0`, the stream will not time out.
    * @returns A Source object that can be passed to a Player to play the audio.
+   * @returns An Emittery instance that emits messages from the WebSocket.
+   * @returns An abort function that can be called to cancel the stream.
    */
-  send(inputs, { timeout = 0 } = {}) {
-    var _a, _b, _c, _d;
+  send(_a, { timeout = 0 } = {}) {
+    var inputs = __objRest(_a, []);
+    var _a2, _b, _c, _d;
     if (!__privateGet(this, _isConnected)) {
       throw new Error("Not connected to WebSocket. Call .connect() first.");
     }
-    const contextId = __privateMethod(this, _generateId, generateId_fn).call(this);
-    (_a = this.socket) == null ? void 0 : _a.send(
-      JSON.stringify(__spreadProps(__spreadValues({
-        context_id: contextId
-      }, inputs), {
-        output_format: {
-          container: "raw",
-          encoding: "pcm_f32le",
-          sample_rate: __privateGet(this, _sampleRate)
-        }
-      }))
+    if (!inputs.context_id) {
+      inputs.context_id = __privateMethod(this, _generateId, generateId_fn).call(this);
+    }
+    if (!inputs.output_format) {
+      inputs.output_format = {
+        container: __privateGet(this, _container),
+        encoding: __privateGet(this, _encoding),
+        sample_rate: __privateGet(this, _sampleRate)
+      };
+    }
+    (_a2 = this.socket) == null ? void 0 : _a2.send(
+      JSON.stringify(__spreadValues({}, inputs))
     );
     const emitter = new Emittery();
     const source = new Source({
-      sampleRate: __privateGet(this, _sampleRate)
+      sampleRate: __privateGet(this, _sampleRate),
+      encoding: __privateGet(this, _encoding),
+      container: __privateGet(this, _container)
     });
     const streamCompleteController = new AbortController();
     let timeoutId = null;
@@ -86,19 +97,26 @@ var WebSocket = class extends Client {
       timeoutId = setTimeout(streamCompleteController.abort, timeout);
     }
     const handleMessage = createMessageHandlerForContextId(
-      contextId,
-      (_0) => __async(this, [_0], function* ({ chunk, message }) {
+      inputs.context_id,
+      (_0) => __async(this, [_0], function* ({ chunk, message, data }) {
         emitter.emit("message", message);
+        if (data.type === "timestamps") {
+          emitter.emit("timestamps", data.word_timestamps);
+          return;
+        }
         if (isSentinel(chunk)) {
           yield source.close();
           streamCompleteController.abort();
           return;
         }
-        yield source.enqueue(base64ToArray([chunk]));
         if (timeoutId) {
           clearTimeout(timeoutId);
           timeoutId = setTimeout(streamCompleteController.abort, timeout);
         }
+        if (!chunk) {
+          return;
+        }
+        yield source.enqueue(base64ToArray([chunk], __privateGet(this, _encoding)));
       })
     );
     (_b = this.socket) == null ? void 0 : _b.addEventListener("message", handleMessage, {
@@ -128,7 +146,11 @@ var WebSocket = class extends Client {
         clearTimeout(timeoutId);
       }
     });
-    return __spreadValues({ source }, getEmitteryCallbacks(emitter));
+    return __spreadProps(__spreadValues({
+      source
+    }, getEmitteryCallbacks(emitter)), {
+      stop: streamCompleteController.abort.bind(streamCompleteController)
+    });
   }
   /**
    * Authenticate and connect to a Cartesia streaming WebSocket.
@@ -198,6 +220,8 @@ var WebSocket = class extends Client {
 };
 _isConnected = new WeakMap();
 _sampleRate = new WeakMap();
+_container = new WeakMap();
+_encoding = new WeakMap();
 _generateId = new WeakSet();
 generateId_fn = function() {
   return humanId({