SwapVoice 2.0.0
Languages: Multilingual
File Size: 1.10 GB
Transform your voice into instant, actionable text with SwapVoice's intelligent trigger recognition system for a onetime subscription fee instead of monthly license with other similar tools. Perfect for professionals, content creators, and anyone who wants to work faster and smarter.
Offline voice recognition
No cloud dependency, complete privacy
Multi-language support
15 languages including English, Dutch, German, French, Spanish, Italian, Portuguese, Ukrainian, Polish, Turkish, Chinese, Japanese, Korean, Arabic, Hindi
Smart trigger system
Say "swap email" to instantly insert your email address
Global hotkey support
Quick access with Ctrl+Alt+V from any application
Voice training
Customize recognition for your specific triggers
Complete offline solution
Works without internet connection
Professional-grade accuracy
Uses advanced OpenAI Whisper technology
Release Notes
Major Engine Upgrade
- Completely rebuilt transcription engine - up to 10x faster than v1.x
- New NVIDIA Canary 1B v2 ONNX engine replaces Python + Whisper
- Native audio capture - no browser encoding overhead
- Audio chunking - long recordings split automatically for fast processing
- Voice Activity Detection - silence trimmed before transcription
- 24 European languages supported (expanded from 15)
- Word correction glossary - fix commonly misspelled names and terms
- VS Code terminal paste support (Ctrl+Shift+V)
- Cancellable transcription - new recording discards in-progress result
- Progress indicator during transcription
- Improved recording popup with processing status
- Security hardened codebase
- Smaller install - no Python runtime needed
- Customizable hotkey (Settings → Hotkey Configuration)
- Improved model loading
System Requirements
- OS: Windows 10 version 17763.0 or higher
- DirectX: Version 9
- Memory: 4 GB
- Video Memory: 1 GB
Quick check before we show the links
Helps us keep automated scrapers from hammering the filehosts.