Legal
Legal notice
Last updated: 4 October 2026
Service provider
[YOUR FULL NAME OR COMPANY NAME]
[STREET AND NUMBER, POSTCODE CITY, COUNTRY]
[OPTIONAL: VAT / TAX ID]
[OPTIONAL: COMPANY REGISTER ENTRY]
Contact
E-mail: [CONTACT EMAIL]
Responsible for content
[YOUR FULL NAME OR COMPANY NAME], address as above.
Affiliate links and donations
Some links on this website may be affiliate links. If you buy something through them, we may receive a commission. The price for you stays the same. Affiliate links are marked as sponsored and do not influence how the tool works. Donations are voluntary and do not unlock any functions.
Trademarks
Product and company names mentioned on this website (e.g. Adobe Podcast, Krisp, Spotify, Apple Podcasts, YouTube, Zoom) are trademarks of their respective owners. SafeAudio Studio is not affiliated with, endorsed or sponsored by them. Comparisons are based on the publicly available information of those products at the time of writing.
Open-source software and AI models
SafeAudio Studio uses the following open-source components. Their license texts are available in the /vendor/licenses/ folder of this website.
| Component | License |
|---|---|
| DeepFilterNet 3 (speech enhancement model and reference implementation) | MIT or Apache-2.0 |
| ONNX Runtime Web | MIT |
| Transformers.js (Hugging Face), used only for transcription | Apache-2.0 |
| Whisper speech-recognition models (OpenAI; ONNX conversions by onnx-community / Xenova), downloaded only when transcription is used | MIT (Whisper); Apache-2.0 / MIT (conversions) |
| lamejs (MP3 encoder) | LGPL-3.0, loaded as a separate, replaceable file |
| JSZip | MIT |
| Inter and JetBrains Mono fonts | SIL Open Font License 1.1 |
| Demo recording (noisy_snr0.wav) from the DeepFilterNet repository | MIT or Apache-2.0 |