Audio is local by default

VoiceTypr runs transcription on your computer by default.

• No audio is uploaded during normal local transcription
• No training-on-your-data language
• No account required for the core offline workflow

Network activity can happen when you choose it:

• Checking for app updates
• Validating your license (if purchased)
• Sending text to an optional AI formatting provider you configured
• Sending audio only if you explicitly use a cloud or network transcription mode

What's Stored Locally

AI Formatting and optional network paths

If you enable AI formatting presets (Default, Prompts, Email, Commit):

• Text may be sent to your chosen AI provider (OpenAI, Anthropic, Google Gemini, or a custom OpenAI-compatible endpoint)
• Audio is not sent for formatting — only the transcribed text
• You control the provider, the key, and whether formatting is enabled

VoiceTypr may also offer optional non-local transcription modes, such as cloud STT or local-network sharing. Those are opt-in paths, not the default offline workflow.

No background telemetry

VoiceTypr does not run background usage analytics.

If you manually submit a support, bug, or crash report, the app can send the details you choose to share, including redacted logs and system information needed to debug the issue.

The website (voicetypr.com) uses privacy-focused analytics (Umami and OpenPanel) with no cookies or cross-site tracking.

Open Source

VoiceTypr is an open-source app built on open-source local transcription models, including Whisper. You can inspect the public desktop repository on GitHub and still keep the normal offline workflow local by default.