What's Changed in v0.6.0-beta
🚀 Major Feature: whisper.cpp is Now Default!
This release introduces whisper.cpp as the new default speech recognition engine — a game-changing upgrade that brings massive improvements to speed, compatibility, and ease of installation.
Performance Improvements
- ⚡ 10x faster installation — No more 2.3GB PyTorch downloads (installation now takes ~1-2 min instead of 5-10 min)
- 🎮 Universal GPU support — Works with AMD, Intel, and NVIDIA via Vulkan (not just NVIDIA CUDA like OpenAI Whisper)
- 💾 Smaller footprint — Tiny model is only ~39MB vs ~75MB for Whisper
- 🔥 Better performance — C++ optimized inference with true multi-threading (no Python GIL)
- 🌍 99+ languages with automatic language detection
✨ New Features
- whisper.cpp integration — High-performance C++ speech recognition with Vulkan GPU acceleration (#179) - @jatinkrmalik
- Interactive installer — Choose between 3 engines: whisper.cpp (recommended), Whisper, or VOSK (#179) - @jatinkrmalik
- Hardware auto-detection — Automatically detects your GPU and recommends optimal settings (#179) - @jatinkrmalik
- Enhanced logging — Comprehensive debug info for whisper.cpp troubleshooting (#179) - @jatinkrmalik
- Cross-distro compatibility — Phase 1-6 complete, supporting Ubuntu, Fedora, Debian, Arch, and more - @jatinkrmalik
- Language indicator — Show detected language with flag in live demo (#169) - @jatinkrmalik
- Auto-detect browser language — For live demo speech recognition (#166) - @jatinkrmalik
🐛 Bug Fixes
- Text injection — Check ydotoold daemon before selecting ydotool on wlroots compositors (#176) - @jatinkrmalik
- Installation — Add python3-dev dependency for evdev wheel build (#177) - @jatinkrmalik
- NVIDIA detection — Prevent silent exit on systems without NVIDIA GPU (#175) - @jatinkrmalik
- Install script — Make quick install command fully visible with text wrap (#156) - @jatinkrmalik
- Website UX — Update install commands to use /tmp/vl.sh for better UX (#154) - @jatinkrmalik
🔧 Infrastructure & CI
- Nightly builds — Improve workflow for user accessibility (#180) - @jatinkrmalik
- Nightly builds — Handle both wheel and tar.gz file renaming - @jatinkrmalik
- CI testing — Actually test install.sh distro detection in CI matrix (#163) - @jatinkrmalik
- Documentation — Add distribution package mapping reference (#162) - @jatinkrmalik
- Screenshots — Add screenshots section to README (#153) - @jatinkrmalik
📚 Documentation Updates
Comprehensive documentation overhaul highlighting whisper.cpp benefits:
- README.md — Major update with whisper.cpp as default messaging - @jatinkrmalik
- docs/INSTALL.md — New installation options and GPU support info - @jatinkrmalik
- docs/USER_GUIDE.md — Engine selection guide added - @jatinkrmalik
- SECURITY.md — Updated supported versions and engine descriptions - @jatinkrmalik
- Website — Cleaner hero layout with tech stack details (#181) - @jatinkrmalik
🎯 Version Bump
- From: v0.5.0-beta
- To: v0.6.0-beta (#181) - @jatinkrmalik
📝 Installation
# Install the latest beta
curl -fsSL https://raw.githubusercontent.com/jatinkrmalik/vocalinux/main/install.sh | bash -s -- --tag=v0.6.0-beta
# Or use interactive mode to choose your engine
curl -fsSL https://raw.githubusercontent.com/jatinkrmalik/vocalinux/main/install.sh | bash -s -- --tag=v0.6.0-beta --interactive📋 Full Commit List
03d55a9 chore(release): prepare v0.6.0-beta with whisper.cpp as default (#181) - Jatin K Malik
318e188 fix(ci): handle both wheel and tar.gz file renaming in nightly builds - jatinkrmalik
4b51e9b fix(ci): improve nightly build workflow for user accessibility (#180) - Jatin K Malik
61b8555 fix(ci): improve nightly build workflow for user accessibility - jatinkrmalik
57a0ddd feat(speech): add whisper.cpp with Vulkan support for GPU acceleration (#179) - Jatin K Malik
4a72c96 fix(text-injection): check ydotoold daemon before selecting ydotool (#176) - Jatin K Malik
5e83362 fix(install): add python3-dev dependency for evdev build (#177) - Jatin K Malik
ff994a3 fix(install): prevent silent exit on systems without NVIDIA GPU (#175) - Jatin K Malik
e9134e0 feat(web): show detected language with flag indicator in live demo (#169) - Jatin K Malik
04eada7 feat: Phase 5 & 6 - Complete cross-distro compatibility fixes (#168) - Jatin K Malik
589d79d fix(web): auto-detect browser language for live demo speech recognition (#166) - Jatin K Malik
8e48b99 ci: Actually test install.sh distro detection in CI matrix (#163) - Jatin K Malik
9efe53e docs: Add distribution package mapping reference (#162) - Jatin K Malik
6b0e81c feat: Phase 4 - Testing & Polish improvements (#161) - Jatin K Malik
461400b feat: Phase 3 - Better error handling and documentation (#160) - Jatin K Malik
f2e13bb feat: Phase 2 - Enhanced distro detection and package manager support (#159) - Jatin K Malik
2b498f0 fix: Phase 1 - Critical cross-distribution compatibility fixes (#158) - Jatin K Malik
3bd9f8e Update README.md - Jatin K Malik
04aa999 fix(web): make quick install command fully visible by enabling text wrap (#156) - Jatin K Malik
257749f fix(website): update install commands to use /tmp/vl.sh for better UX (#154) - Jatin K Malik
03caf47 feat(readme): add screenshots section to README (#153) - Jatin K Malik
🙏 Contributors
Thanks to everyone who contributed to this release!
Full Changelog: v0.5.0-beta...v0.6.0-beta