VoiceFlow v0.3.0 — Smart Voice Actions, Dynamic Domain Launcher, Adaptive Silence & Hardware Push-to-Talk

VoiceFlow v0.3.0 — The Zero-Latency Voice Operating System ⚡

Type at the speed of thought. 100% offline, private AI voice dictation and system-wide voice actions for Windows and macOS.

VoiceFlow v0.3.0 is our biggest release yet. It transforms VoiceFlow from a pure speech-to-text dictation utility into a complete voice operating system launcher, featuring Smart Voice Actions, Dynamic Domain Resolution, Adaptive Silence Auto-Submit (Zero-Enter), and Hardware Push-to-Talk.

---

🌟 What's New in v0.3.0

1. 🚀 Smart Voice Actions & Universal Launcher ("Open Command")

Speak naturally to summon apps, open websites, and trigger web searches with zero latency:

• Native App Launcher: Say "open VS Code", "open Spotify", "open Notepad", "open Terminal", or "open Calculator". VoiceFlow intercepts the spoken intent, flashes a sleek purple action badge on the HUD, and launches the app directly—without typing a single character into your active document.
• Dynamic Domain Resolver: Say "open gdha.online", "open myportfolio.dev", or "open github.com". VoiceFlow dynamically recognizes 20+ top-level domains (.online, .dev, .ai, .io, .app, .tech, .org, etc.), formats a clean HTTPS URL, and launches it in your default browser.
• Smart Web Search: Say "search YouTube for lo-fi beats" or "search Google for rust tauri shortcuts" to route queries directly to the provider's native search engine.

2. ⏱️ Adaptive Silence Auto-Submit (Zero-Enter Dictation)

No more pressing Enter to submit your voice commands or dictation:

• Command Fast-Track (750ms): Spoken voice actions ("open...", "search...") automatically detect trailing silence and execute in just 750ms for instantaneous, hands-free automation.
• Forgiving Dictation Window (2,200ms): For long emails, Slack messages, or documentation, VoiceFlow provides a 2.2-second silence cushion. At 1.2s of silence, a subtle countdown badge (1s) appears. If you resume speaking mid-pause, the timer immediately resets and dictation continues seamlessly.

3. 🎙️ Hardware Push-to-Talk (Ctrl + Space)

Engineered with an elevated kernel-level Win32 low-level keyboard hook (WH_KEYBOARD_LL) running at time-critical scheduling priority:

• Hold to Speak & Release to Submit: Hold Ctrl + Space, speak your thoughts, and release. The microsecond either key is released after $\ge 350\text{ms}$, VoiceFlow instantly commits and pastes your transcript.
• Quick Tap ($<350\text{ms}$): Tap Ctrl + Space quickly to keep the HUD open for continuous hands-free dictation.
• Document Cleanliness: Consumes keydown events at the OS level so space characters never leak into your code editor or document.

4. 🗣️ Hands-Free Offline Wake Word ("Hey Flow")

• 100% Offline Matching: Native MFCC feature extraction + Dynamic Time Warping (DTW) pattern matcher.
• Ultra-Low Idle CPU (<1%): Adaptive VAD noise-floor gating prevents background battery drain.
• One-Breath Pre-Roll PCM Buffer: 300ms circular ring buffer ensures words spoken in the very same breath ("Hey Flow open VS Code") are never clipped.
• Personalized Voice Profile: 3-step rapid voice calibration with medoid clustering tunes the detector specifically to your vocal pitch and timbre.

---

🛡️ Privacy & Performance Guarantee

• 100% Local Inference: Zero audio or text bytes ever leave your device.
• Air-Gapped & Offline: Works seamlessly in Airplane Mode without an internet connection.
• Ultra-Light Base Footprint: ~34 MB installer size with zero heavy bundled weights.

VoiceFlow v0.2.5

What's New in v0.2.5

• Signature Coral Brand Theme: 100% unified Coral #FF5722 design tokens across all views and buttons (zero blue).
• Overlay Window Transparency Fix: Eliminated the rectangular border layer on the Ctrl + Space floating capsule in Windows 11.
• New App Icons: High-resolution Coral icon for taskbar, title bar, and system tray.
• Custom Vocabulary Packs: 1-click domain preset terminology packs (Developer, AI/ML, SaaS) + Whisper acoustic tuning guide.
• Performance: Standalone self-contained backend bundle.

VoiceFlow v0.2.4

Initial commit

VoiceFlow v0.2.1

Full Changelog: https://github.com/Utpal4505/voiceflow-releases/commits/v0.2.2

VoiceFlow v0.2.1

VoiceFlow v0.2.0 - Standalone engine bundle with zero external prerequisites.

Full Changelog: https://github.com/Utpal4505/voiceflow-releases/commits/v0.2.1

VoiceFlow v0.2.0

Full Changelog: https://github.com/Utpal4505/voiceflow-releases/commits/v0.2.0