v0.13.0-beta
BetaLatest2026-06-30- Guided whisper.cpp model selection: pick a size plus a specialization (English-only, quantized Q5/Q8, or Large v3 Turbo) through split Model Size and Specialization dropdowns with in-app guidance; the --model flag also accepts exact IDs like medium.en-q5_0 and large-v3-turbo (PR #465)
- Dictation now keeps a space between segments spoken with a pause in between, so words no longer run together after a silence (PR #464)
- Keyboard shortcuts now work on keyboards hotplugged after Vocalinux starts, with the evdev backend rescanning for new devices and recovering from disconnects (PR #467)
- Vocalinux now detects KDE Plasma Wayland sessions and points you to enable IBus Wayland for reliable text injection, surfaced during install and when wtype injection fails (PR #466)
- Wayland: fixed garbled text on non-US keyboard layouts (AZERTY/QWERTZ/Dvorak) and a clipboard-copy hang; ydotool now pastes through the clipboard, which is layout-independent (PR #480)
- Wayland: use wtype/ydotool instead of IBus on compositors that don't bridge it to native apps like COSMIC, Sway, and Hyprland, fixing silent text drops (PR #486)
- Wayland/IBus: require a real IM engine before using IBus, so a bare xkb layout no longer causes silent text drops on GNOME/Mutter and other compositors (#478)
- Wayland: keep the keyboard layout intact by not running setxkbmap, which was flipping XWayland apps to us after dictation (#474)
- Faster ydotool text injection via an explicit --key-delay (PR #488)
- Settings dialog height capped on high-resolution displays (PR #465)
- Refreshed website docs with new feature pages for Remote API, Silero VAD, advanced whisper.cpp settings, and desktop reliability, plus responsive layout polish (PR #470)