CREATE AI VOICEOVERS — SECOND DESIGN VARIANT Font: DM Sans, SIL Open Font License (assets/DM-Sans-OFL.txt). Inference: ONNX Runtime Web 1.18.0, Microsoft, MIT License. https://github.com/microsoft/onnxruntime/tree/v1.18.0 The local browser bundle adds only an ES-module export to the upstream UMD bundle. Full license: onnxruntime-MIT.txt Phonemizer JS wrapper: @diffusionstudio/vits-web 1.0.3, package metadata: MIT. https://github.com/diffusionstudio/vits-web The included phonemizer module is unmodified. Phonemizer WebAssembly/data: @diffusionstudio/piper-wasm 1.0.0. https://www.npmjs.com/package/@diffusionstudio/piper-wasm https://github.com/diffusionstudio/piper-wasm Includes eSpeak NG, GPL version 3 or later, and Piper phonemizer components. Corresponding eSpeak NG source: https://github.com/espeak-ng/espeak-ng License: espeak-GPL-3.txt Voice models are fetched directly from: https://huggingface.co/diffusionstudio/piper-voices https://huggingface.co/rhasspy/piper-voices Models are not bundled with this static site. Model cards and original dataset attributions are included in voice-models/. Supported voices use datasets identified by those cards as public domain/CC0 or CC BY 4.0. Alba: The Centre for Speech Technology Research, University of Edinburgh, https://datashare.ed.ac.uk/handle/10283/3270 , CC BY 4.0. Siwis: SIWIS French Speech Synthesis Database, University of Edinburgh, https://datashare.is.ed.ac.uk/handle/10283/2353 , CC BY 4.0. Attribution/license details are preserved in each corresponding model card. Voice-character photographs: Unsplash License. Individual authors and source URLs: portrait-photos.txt