Orca Streaming Text-to-Speech
Web Quick Start

Platforms

• Chrome & Chromium-based browsers
• Edge
• Firefox
• Safari

Requirements

• Picovoice Account and AccessKey
• Node.js 16+
• npm

Picovoice Account & AccessKey

Signup or Login to Picovoice Console to get your AccessKey. Make sure to keep your AccessKey secret.

Quick Start

Setup

1. Install Node.js.

2. Install the Orca Streaming Text-to-Speech Web package:

npm install @picovoice/orca-web

Usage

Orca Streaming Text-to-Speech can synthesize speech in different languages and with a variety of voices, each of which is characterized by a model file (.pv). Download the model file for the language/voice that you prefer. The language and gender of the speaker is indicated in the file name.

Put the model file in the project's public directory or generate a base64 model using the built-in script:

npx pvbase64 -i ${ORCA_PARAMS_PATH} -o ${OUTPUT_FILE_PATH}

Create an OrcaWorker instance using a base64 model or a model hosted in a public directory:

import { OrcaWorker } from "@picovoice/orca-web";
import orcaParams from "${ORCA_PARAMS_BASE64_PATH}";

const orca = await OrcaWorker.create(
  "${ACCESS_KEY}",
  {
    base64: orcaParams,
    // or
    publicPath: "${MODEL_RELATIVE_PATH}",
  }
);

Orca Streaming Text-to-Speech supports two modes of operation: streaming and single synthesis. In the streaming synthesis mode, Orca processes an incoming text stream in real-time and generates audio in parallel. In the single synthesis mode, a complete text is synthesized in a single call to the Orca engine.

Streaming synthesis

To use streaming synthesis, call streamOpen to create an OrcaStream object.

const orcaStream = await orca.streamOpen();

Then, call synthesize on orcaStream to generate speech from a stream of text:

function textGenerator(): string[] {
  // yield text chunks e.g. tokens from an LLM response
}

for (const textChunk of textGenerator()) {
  const pcm = await orcaStream.synthesize(textChunk);
  if (pcm !== null) {
    // handle pcm
  }
}

The textGenerator() function can be any stream generating text, such as an LLM response.

The OrcaStream object buffers input text until there is enough context to generate audio. If there is not enough text to generate audio, null is returned.

Once the text stream is complete, call the flush method to synthesize the remaining text:

const flushedPcm = orcaStream.flush();
if (flushedPcm !== null) {
  // handle pcm
}

When done with streaming text synthesis, the OrcaStream object needs to be closed:

orcaStream.close();

Single synthesis

Synthesize speech by calling the synthesize method:

// returns raw PCM and alignments
const { pcm, alignments } = await orca.synthesize("${TEXT}");

Release resources

Release resources explicitly when done with Orca:

await orca.release()

Demo

For the Orca Streaming Text-to-Speech Web SDK, there is a Web demo project available on the Orca GitHub repository.

Setup

Clone the Orca Streaming Text-to-Speech repository from GitHub:

git clone --recurse-submodules https://github.com/Picovoice/orca.git

Usage

1. Install dependencies and run:

cd orca/demo/web
npm install
npm run start

2. Open http://localhost:5000 to view it in the browser.

Resources