Model & Open-Source Licenses

The on-device models are downloaded from Hugging Face on first use and cached locally; the runtimes ship inside the extension. By using Rongo's on-device voices and on-device language model, you agree to the use restrictions of the Supertonic and Gemma models reproduced below.

On-device models

Model	Used for	License
Supertonic 3	Polish TTS (Supertonic engine)	BigScience Open RAIL-M (weights); MIT (inference code)
Gemma 4 E2B	On-device Summarise / Translate	Gemma Terms of Use
Kokoro-82M	English TTS	Apache-2.0
Piper voices	Polish TTS (Piper engine)	MIT (repo); per-voice dataset licenses
Whisper (tiny.en / base)	Optional karaoke word-timing	MIT

1. Supertonic 3 — Open RAIL-M (use restrictions apply)

Polish on-device voices. Source: Supertone/supertonic-3. Author: Supertone, Inc. The model weights are licensed under the BigScience Open RAIL-M License (the ported inference code is MIT). Commercial use is permitted, but the license imposes use-based restrictions that pass to every downstream user.

By using the Supertonic voices you agree NOT to use the model, or allow it to be used: to break any applicable law; to exploit or harm minors; to generate or spread verifiably false information to harm others; to generate or spread personal data to harm someone; to defame or harass; for fully automated decisions affecting someone's legal rights; to discriminate against or harm people based on social behaviour or personal characteristics; to exploit the vulnerabilities of a specific group; to discriminate based on legally protected characteristics; to provide medical advice or interpret medical results; or to generate information for justice / law-enforcement / immigration / asylum processes (e.g. predicting that an individual will commit a crime). The full license is in the model repository.

2. Gemma 4 E2B — Gemma Terms of Use

On-device Summarise / Translate model. Source: onnx-community/gemma-4-E2B-it-ONNX. Author: Google LLC (ONNX conversion by ONNX Community).

Required notice: “Gemma is provided under and subject to the Gemma Terms of Use found at ai.google.dev/gemma/terms.” Use of the model is governed by the Gemma Terms of Use and the Gemma Prohibited Use Policy. By using Rongo's on-device Summarise / Translate you agree to those terms.

3. Kokoro-82M — Apache-2.0

English on-device voices. Source: onnx-community/Kokoro-82M-v1.0-ONNX. Author: hexgrad (Kokoro-82M); ONNX conversion by ONNX Community; built on StyleTTS 2 (yl4579/StyleTTS2-LJSpeech). Licensed under the Apache License 2.0.

4. Piper voices — MIT

Polish on-device voices for the Piper engine, from rhasspy/piper-voices (MIT; mirrored at diffusionstudio/piper-voices). Voices shipped: gosia, darkman, mc_speech, mls_6892. Each voice's training-data license is recorded in its per-voice MODEL_CARD (licenses vary by source dataset — e.g. Multilingual LibriSpeech for the mls_* voices).

Community Polish voices jarvis and justyna come from WitoldG/polish_piper_models (declared MIT; voice-source provenance undocumented).

5. Whisper — MIT

Optional on-device karaoke word-timing. Source: onnx-community/whisper-tiny.en_timestamped and onnx-community/whisper-base_timestamped. Author: OpenAI (Whisper); ONNX conversion by ONNX Community. Licensed under the MIT License.

Bundled / loaded runtimes

Component	Author	License
ONNX Runtime Web	Microsoft	MIT
Transformers.js	Hugging Face	Apache-2.0
kokoro-js	hexgrad / Xenova	Apache-2.0
piper-tts-web	Mintplex Labs	MIT
piper-phonemize / espeak-ng	espeak-ng	GPL-3.0 (loaded from CDN, not redistributed in-package)

Corrections

Questions or corrections about attribution: meethemoser@gmail.com.