Credits & Licenses

FreeVoice Reader is built on the work of many open-source projects. This page lists the software and models we redistribute or rely on, along with their license terms and source links.

eSpeak NG

GPL-3.0-or-later

Compact open-source software speech synthesizer that converts text to phonemes for the Kokoro TTS pipeline running in your browser. Bundled as a Web Worker (espeakng.worker.js) and served alongside Kokoro assets.

Source code (GitHub)License (GPL-3.0)

Kokoro TTS (model)

Apache-2.0

82M-parameter open-weight text-to-speech model used for in-browser local synthesis. Weights are fetched from Hugging Face the first time you use Local mode.

Model (Hugging Face)Project (GitHub)License (Apache-2.0)

kokoro-js

MIT

JavaScript runtime that drives Kokoro inference in the browser via transformers.js and ONNX Runtime Web.

Source code (GitHub)License (MIT)

Transformers.js

Apache-2.0

Hugging Face library for running machine-learning models directly in the browser, with WebGPU and WebAssembly back-ends.

Source code (GitHub)Documentation License (Apache-2.0)

ONNX Runtime Web

MIT

Microsoft's cross-platform inference engine, compiled to WebAssembly and used to execute Kokoro on-device. The runtime WASM binaries are served from this site for CSP compatibility.

Source code (GitHub)License (MIT)

Supertonic TTS

See repository

Open-source local TTS with flow-matching synthesis, used as an alternative to Kokoro in the browser-side voice picker.

Model (Hugging Face)Source code (GitHub)Paper (arXiv)

Microsoft Edge Read-Aloud TTS

Microsoft service terms

Free Cloud TTS tier uses Microsoft Edge's public read-aloud neural voices. This is a third-party hosted service, not open source.

Azure Text-to-Speech

Google Cloud Text-to-Speech

Google Cloud service terms

Pro and Premium tiers use Google Cloud TTS for Neural2, Studio, and Chirp voices. Hosted third-party service, not open source.

Google Cloud TTS

Next.js, React, Tailwind CSS, and others

Various (MIT, Apache-2.0, etc.)

FreeVoice Reader is built on top of the broader open-source web ecosystem, including Next.js, React, Tailwind CSS, Drizzle ORM, Radix UI, and many more libraries. We're grateful to all of their authors and contributors.

Notice for redistributed components

The Kokoro Local TTS feature ships an unmodified copy of the eSpeak NG phonemizer Web Worker as part of the browser bundle. eSpeak NG is licensed under the GNU General Public License version 3 or later, and its complete corresponding source code is available at github.com/espeak-ng/espeak-ng. No part of FreeVoice Reader's proprietary code is statically linked into eSpeak NG.

Notice an omission or want us to fix attribution? Let us know.