Waveforge — Privacy Policy

Waveforge is built by 3Nsofts (Ehsan Azish). This policy explains, in plain language, what Waveforge does and does not do with your data. The short version: Waveforge does not collect your audio or upload it for processing. You control whether to export files to another app or service.

The short version

• Recording, editing and export work on-device. Optional transcription models require a download before local use.
• There are no accounts, no sign-in, no sign-up.
• There is no analytics, no telemetry, no tracking, no advertising.
• Your audio, recordings, transcripts, and projects are not uploaded to us. You can choose to export through Files or Share.
• Audio recording, editing and transcription run on your device.

What Waveforge processes

Waveforge works with the audio you import or record: voice recordings, music files, and the projects you build from them. All of this — your audio files, mixes, effects, transcripts, and saved .waveforge projects — is processed and stored locally on your device. None of it is sent to us for processing. Files providers and other apps receive only the content you choose to export or share.

Transcription and on-device AI

Waveforge's transcription feature runs entirely on your device using a speech-recognition model that you optionally download once. Your audio is transcribed locally on your device. No audio or transcript is sent to a server for recognition. If you choose not to install the transcription model, that feature is simply unavailable — and the rest of the app works without it.

The one-time model download fetches the speech-recognition model from its public distribution source. The model host may receive ordinary network request information such as your IP address; the download does not include your audio or transcript and happens when you explicitly choose to enable transcription.

What we collect

Nothing. Waveforge does not collect any personal information, usage data, diagnostics, or analytics. We have no servers that receive your data because Waveforge never sends any.

Microphone access

Waveforge requests microphone access only when you choose recording or voiceover, so it can capture your audio. Recordings are saved locally to your project. Your microphone audio is not transmitted to us. You can revoke microphone access at any time in your device’s Settings → Privacy & Security → Microphone.

File access

Waveforge accesses only the audio files and projects you explicitly open, import, or drop into the app, using Apple’s standard sandboxed file access on Mac, iPhone and iPad. On iOS the app keeps local working copies of imported audio and projects. It cannot and does not scan, index, or access files you have not chosen to give it.

Local storage

Waveforge stores your projects, recordings, and (if enabled) the downloaded transcription model in standard local locations on your device. Saved .waveforge projects are self-contained packages that include copies of your audio so your projects remain intact and portable. Removing the iOS app deletes its local app data. Copies exported to Files, cloud providers or other apps remain until you delete them there. On Mac, exported projects also remain separate from the app.

Third parties

Optional speech-model downloads use WhisperKit and public Hugging Face model hosting. Recognition runs locally. Files locations (including iCloud Drive) and destinations you choose through Share handle your exports under their own policies. We do not use third-party analytics, advertising, or tracking SDKs.

Children's privacy

Waveforge does not collect any data from anyone, including children. It is safe to use at any age in this respect.

Changes to this policy

If this policy changes, the updated version will be posted here with a revised "Last updated" date. Because Waveforge collects no data, material changes are unlikely.

Contact

Questions about this policy: [email protected]

Waveforge is available for Mac, iPhone and iPad. Audio processing stays on-device; network access is used for optional models and user-selected file/share services.