Audio retention

Audio files are the most sensitive thing Spellar handles. This page is the specific set of controls and defaults around them.

Defaults

• Local-only recordings: audio never leaves your device. There's no retention from us because we never have it.
• Server-transcribed recordings: audio is stored in encrypted R2 storage, retained until you delete the recording or opt out per-recording.

Per-recording opt-out (3.0.2+)

Every recording has an Audio Privacy setting. You can choose to delete the audio immediately after transcription, keeping only the transcript and summary.

1. Open a recording → menu → Audio Privacy → 'Delete audio after processing'
2. Once transcription completes, the audio file is purged from R2 within 24 hours

<aside> 🖼️ 📷 [screenshot: Audio Privacy setting in recording menu]

</aside>

Workspace defaults (Pro+)

Workspace admins can set a default: 'always retain audio', 'always purge after processing', or 'user choice per recording'. Settings → Workspace → Privacy.

Manual deletion

• Delete a recording → audio is queued for deletion, purged from R2 within 7 days
• Delete account → all audio cascades within 30 days, including backups (rolling)
• Bulk-delete all recordings — Settings → Account → Bulk Delete

What stays after audio deletion

• Transcript (unless you also delete the recording)
• Summary (unless you also delete the recording)