Kokoro TTS

https://www.npmjs.com/package/kokoro-js
Admin
Kokoro TTS is a JavaScript library that enables the Kokoro 82 million parameter text-to-speech (TTS)...
Tokens:5,547
Snippets:33
Trust Score:9.6
Update:1 month ago
Show doc for...
Tokens:
Raw
### Install kokoro-js

Source: https://www.npmjs.com/package/kokoro-js?activeTab=dependencies

Command to install the library via npm.

```bash
npm i kokoro-js
```

--------------------------------

### KokoroTTS.from_pretrained

Source: https://www.npmjs.com/package/kokoro-js?activeTab=code

Initializes the Kokoro TTS model from a specified model ID with configuration options for data type and compute device.

```APIDOC
## KokoroTTS.from_pretrained(model_id, options)

### Description
Loads the Kokoro TTS model. This is an asynchronous operation that returns a KokoroTTS instance.

### Parameters
- **model_id** (string) - Required - The Hugging Face model ID (e.g., "onnx-community/Kokoro-82M-v1.0-ONNX").
- **options** (object) - Optional
  - **dtype** (string) - Optional - Data type for the model. Options: "fp32", "fp16", "q8", "q4", "q4f16".
  - **device** (string) - Optional - Compute device. Options: "wasm", "webgpu" (web) or "cpu" (node).
```

--------------------------------

### KokoroTTS.from_pretrained

Source: https://www.npmjs.com/package/kokoro-js?activeTab=versions

Initializes the Kokoro TTS model from a specified model ID with configuration options for data type and device.

```APIDOC
## KokoroTTS.from_pretrained(model_id, options)

### Description
Loads the Kokoro TTS model from a Hugging Face model repository.

### Parameters
- **model_id** (string) - Required - The identifier of the model to load (e.g., "onnx-community/Kokoro-82M-v1.0-ONNX").
- **options** (object) - Optional
  - **dtype** (string) - Optional - Data type for the model. Options: "fp32", "fp16", "q8", "q4", "q4f16".
  - **device** (string) - Optional - Execution device. Options: "wasm", "webgpu" (web), or "cpu" (node).
```

--------------------------------

### KokoroTTS.from_pretrained

Source: https://www.npmjs.com/package/kokoro-js?activeTab=dependencies

Initializes the Kokoro TTS model from a specified model ID with configuration options for data type and device.

```APIDOC
## KokoroTTS.from_pretrained(model_id, options)

### Description
Loads a pre-trained Kokoro TTS model.

### Parameters
- **model_id** (string) - Required - The Hugging Face model ID.
- **options** (object) - Required
  - **dtype** (string) - Optional - Data type: "fp32", "fp16", "q8", "q4", "q4f16".
  - **device** (string) - Optional - Execution device: "wasm", "webgpu" (web) or "cpu" (node).
```

--------------------------------

### Run End-to-End Tests

Source: https://www.npmjs.com/package/kokoro-js-jp?activeTab=dependencies

Commands to prepare the environment and execute end-to-end tests using Playwright. Note that tests target the dist/ directory, requiring a build before execution.

```bash
npm run build          # test/e2eはdist/を対象にする(src/を直接は見ない)。変更後は必ず再ビルド
npx playwright install chromium   # 初回のみ
npm run test:e2e
```

--------------------------------

### KokoroTTS.from_pretrained

Source: https://www.npmjs.com/package/kokoro-js

Initializes and loads the Kokoro TTS model from a specified model ID.

```APIDOC
## KokoroTTS.from_pretrained(model_id, options)

### Description
Loads the TTS model from the specified Hugging Face model repository.

### Parameters
- **model_id** (string) - Required - The identifier of the model to load (e.g., "onnx-community/Kokoro-82M-v1.0-ONNX").
- **options** (object) - Required
  - **dtype** (string) - Optional - Data type for the model weights. Options: "fp32", "fp16", "q8", "q4", "q4f16".
  - **device** (string) - Optional - Execution device. Options: "wasm", "webgpu" (browser), or "cpu" (node).
```

--------------------------------

### Generate speech with KokoroTTS

Source: https://www.npmjs.com/package/kokoro-js?activeTab=dependencies

Basic usage to initialize the model and generate an audio file from text.

```javascript
import { KokoroTTS } from "kokoro-js";

const model_id = "onnx-community/Kokoro-82M-v1.0-ONNX";
const tts = await KokoroTTS.from_pretrained(model_id, {
  dtype: "q8", // Options: "fp32", "fp16", "q8", "q4", "q4f16"
  device: "wasm", // Options: "wasm", "webgpu" (web) or "cpu" (node). If using "webgpu", we recommend using dtype="fp32".
});

const text = "Life is like a box of chocolates. You never know what you're gonna get.";
const audio = await tts.generate(text, {
  // Use `tts.list_voices()` to list all available voices
  voice: "af_heart",
});
audio.save("audio.wav");
```

--------------------------------

### KokoroTTS.stream

Source: https://www.npmjs.com/package/kokoro-js?activeTab=code

Initializes a streaming synthesis session using a TextSplitterStream.

```APIDOC
## KokoroTTS.stream(splitter)

### Description
Returns an async generator that yields audio chunks as text is pushed into the provided splitter.

### Parameters
- **splitter** (TextSplitterStream) - Required - An instance of TextSplitterStream used to feed text into the model.
```

--------------------------------

### tts.stream

Source: https://www.npmjs.com/package/kokoro-js?activeTab=dependencies

Initializes a streaming interface for generating audio from text chunks.

```APIDOC
## tts.stream(splitter)

### Description
Returns an async iterable stream for processing text chunks.

### Parameters
- **splitter** (TextSplitterStream) - Required - The stream splitter instance.
```

--------------------------------

### tts.generate

Source: https://www.npmjs.com/package/kokoro-js?__cf_chl_rt_tk=o5Yx4uPCvz0z0RBJ6roKoyW3B1MVp8MX9lo9vVQMSm8-1788467770-1.0.1.1-Uf0xrHIVzcHdyRFhzWSJv5QsqZ2LA1Hu5Jt7TrLZFUg

Generates speech audio from the provided text.

```APIDOC
## tts.generate(text, options)

### Description
Generates audio from input text using the loaded model.

### Parameters
- **text** (string) - Required - The text to convert to speech.
- **options** (object) - Optional
  - **voice** (string) - Required - The identifier for the voice to use (e.g., "af_heart").
```

--------------------------------

### KokoroTTS.generate

Source: https://www.npmjs.com/package/kokoro-js?activeTab=code

Generates speech audio from the provided text using a specified voice.

```APIDOC
## KokoroTTS.generate(text, options)

### Description
Generates audio from input text. Returns an audio object that can be saved or processed.

### Parameters
- **text** (string) - Required - The text to synthesize.
- **options** (object) - Optional
  - **voice** (string) - Required - The identifier for the voice to use (e.g., "af_heart").
```

--------------------------------

### tts.generate

Source: https://www.npmjs.com/package/kokoro-js?activeTab=dependencies

Generates audio from the provided text using a specific voice.

```APIDOC
## tts.generate(text, options)

### Description
Generates speech audio from input text.

### Parameters
- **text** (string) - Required - The text to convert to speech.
- **options** (object) - Required
  - **voice** (string) - Required - The identifier for the voice to use.
```

--------------------------------

### Stream speech output with TextSplitterStream

Source: https://www.npmjs.com/package/kokoro-js?activeTab=dependencies

Advanced usage to stream text input and generate audio chunks incrementally.

```javascript
import { KokoroTTS, TextSplitterStream } from "kokoro-js";

const model_id = "onnx-community/Kokoro-82M-v1.0-ONNX";
const tts = await KokoroTTS.from_pretrained(model_id, {
  dtype: "fp32", // Options: "fp32", "fp16", "q8", "q4", "q4f16"
  // device: "webgpu", // Options: "wasm", "webgpu" (web) or "cpu" (node).
});

// First, set up the stream
const splitter = new TextSplitterStream();
const stream = tts.stream(splitter);
(async () => {
  let i = 0;
  for await (const { text, phonemes, audio } of stream) {
    console.log({ text, phonemes });
    audio.save(`audio-${i++}.wav`);
  }
})();

// Next, add text to the stream. Note that the text can be added at different times.
// For this example, let's pretend we're consuming text from an LLM, one word at a time.
const text = "Kokoro is an open-weight TTS model with 82 million parameters. Despite its lightweight architecture, it delivers comparable quality to larger models while being significantly faster and more cost-efficient. With Apache-licensed weights, Kokoro can be deployed anywhere from production environments to personal projects. It can even run 100% locally in your browser, powered by Transformers.js!";
const tokens = text.match(/\s*\S+/g);
for (const token of tokens) {
  splitter.push(token);
  await new Promise((resolve) => setTimeout(resolve, 10));
}

// Finally, close the stream to signal that no more text will be added.
splitter.close();

// Alternatively, if you'd like to keep the stream open, but flush any remaining text, you can use the `flush` method.
// splitter.flush();
```
Tokens:
Raw