Privacy Policy — Zero Audio Retention
How DictateBar (dictatebar.com) protects your voice data with browser-local Web Speech API and Zero Audio Retention cloud streaming.
Last updated: 2026-11-16
1. Overview & Zero Audio Retention Commitment
DictateBar (dictatebar.com) is designed around a strict Zero Audio Retention architecture:
- Free Browser-Local Mode: Real-time voice-into-text runs locally inside your browser using the native Web Speech API. DictateBar servers receive 0 bytes of audio, and no transcripts are stored on our servers.
- Pro Cloud Mode (Whisper Large v3 Turbo & LLM Smart Polish): When you explicitly invoke cloud noise-resistant transcription or AI Smart Polish, audio or text is processed strictly in transient memory (RAM) for the duration of the request and discarded immediately after returning the response.
- No AI Model Training: We never store your audio recordings on disk and never use your voice or transcript data to train, fine-tune, or improve third-party or internal AI models.
2. Information We Collect
- Account Information: If you create an account, we store your email address and authentication metadata via Better-Auth.
- Billing Information: Payments are processed securely through Waffo Pancake. DictateBar does not store raw credit card numbers on its servers.
- Optional User Preferences: Personal dictionary terms or voice macros you explicitly save to your account.
3. Data Security & Third-Party Processors
We rely on Cloudflare Workers and Cloudflare D1 for application hosting, and ephemeral inference APIs solely to fulfill explicit Pro transcription/polish requests. All traffic is encrypted in transit via TLS.
4. Your Rights & Contact
You may request account deletion, data export, or privacy assistance at any time by emailing support@dictatebar.com (response SLA: under 24 hours).