Terms of use

Plain-language terms for using Cadenza, version 2026-10-03. They are not legal advice. 简体中文

1. The software

Cadenza is free, open-source software released under the MIT License (see LICENSE). You may use, copy, modify and share it under that license.

2. No warranty

The software is provided “as is”, without warranty of any kind. Speech recognition can be wrong, incomplete or slow, and text may not be inserted where you expect. Check what is written before you send, save or rely on it. Do not rely on it where an error could cause harm, such as medical, legal, financial or safety-critical work.

3. Your responsibilities

Record only when you are allowed to. If a recording includes other people, make sure you have their permission where the law or common courtesy requires it.
Cloud recognition uses your own account with a speech provider. Their terms, prices, quotas and data policies apply to you. Turning on upload for a provider is your decision, made separately for each provider.
Keep your provider keys and the local API access token private. Anyone with the token can use the local API on your Mac.
Do not use Cadenza to break the law or to harm others.

4. Third-party services and models

Speech providers, downloadable models and their trademarks belong to their owners. Cadenza is not affiliated with or endorsed by them. Models have their own licenses, shown in Settings → Engines → Local. Read them before you use or share a model.

5. Privacy

How your data is handled is described in the Privacy notice. In short: Cadenza collects nothing, and your audio goes only where you choose to send it.

6. Changes

These terms and the privacy notice carry a version date. When they change in a way you should re-read, the app asks you to review them again. Continuing to use the app after you accept a new version means you accept it.

7. Questions and problems

Use the project’s public issue tracker. Do not post recordings, private text, credentials or unredacted logs there.