Vowen v0.1.16 — Release Notes

New Features

Live Real-Time Transcription

See your words appear as you speak. Vowen now supports real-time transcription for Soniox and all applicable cloud models, plus a live transcription preview for Parakeet. Models that support real-time streaming are marked with a "Real-time capable" pill in the model selector.

Tones Mode

Control the tone and style of your transcriptions with a brand new Tones mode. Choose between manual and auto-detection to shape how your voice is transcribed — complete with style examples and few-shot prompts. An indicator shows which modes have tones enabled.

File Tagging for Cursor & Windsurf

Tag files directly from your voice while coding. Enable file tagging in settings to reference project files hands-free in Cursor and Windsurf, with the option to add folders as well.

Speaker Diarization

Identify and label different speakers in Meeting Notes and Manual Transcription modes. Available for models that support diarization, making multi-speaker recordings much easier to follow.

Keychain Encryption for API Keys

API keys are now securely stored using your system's native keychain (macOS Keychain / Windows Credential Manager) providing stronger protection for your credentials.

Azure OpenAI Support

Connect to Azure OpenAI compatible endpoints for AI enhancement and transcription. You can also set custom display names for any Custom API model.

Improvements

Redesigned Recording Indicator

The recording indicator has been completely redesigned with an improved audio waveform visualizer and now displays the currently active app alongside your recording status.

Cancel Hands-Free Mode with Esc

Quickly exit hands-free mode by pressing the Escape key — no need to reach for the shortcut.

Junk Transcription Filtering

Vowen now automatically filters out noise and filler like "thank you" and empty spaces, so only meaningful transcriptions come through.

Tooltips & UI Polish

Added helpful tooltips across the app, refreshed menu items and model icons, improved voice log formatting with properly formatted entries, and standardized log containers with an increased window size.

More Groq Models

Expanded the list of available Groq models for even more transcription options, with reduced latency in Tones mode.

Onboarding Improvements

Added a restart button to the onboarding flow, keys now light up during the keyboard shortcut step, and keyspy starts automatically after the permissions step.

Fixes

Windows Hands-Free Mode

Resolved an issue with hands-free mode getting stuck on Windows, plus general toolbar and active-app detection fixes for Windows users.

Stuck Recording State

Fixed an issue where the recording state could get stuck by implementing a 2-way acknowledgment protocol between the main and renderer processes.

Other Changes

• Open at login support for Windows

• Dictionary is now sorted A-Z with duplicates removed

• Screen recording shortcut removed in favor of UI-only control

• Added feature flags system for experimental features

📥 Download & Update

If you haven't already, Vowen will automatically notify you about this update. You can also:

Auto-update: Accept the prompt when it appears

Manual download: Visit vowen.ai

Thank You 🙏

We appreciate your continued support and feedback! This release is packed with highly requested features — live transcription, tones, file tagging, and speaker diarization — alongside major security and UX improvements.

Have questions, feature requests, or found a bug? We'd love to hear from you:

Email: hello@vowen.ai

Changelog: vowen.featurebase.app/changelog

Feature Requests: vowen.featurebase.app