Repository navigation
v0.7.0 - the voice & TUI release
KrillLM v0.7.0 brings a full-screen chat TUI and on-device voice to krillm run.
Highlights
- Full-screen chat TUI (default on a TTY): alternate-screen interface with a scrollable conversation pane, slash-command autosuggest, streaming markdown, and resize-aware layout.
--classicfor the line REPL; auto-fallback when stdout is not a TTY. - Interactive multimodal input: attach images / audio / mic clips inline.
- On-device dictation: push-to-dictate via Apple Speech plus a native MLX Whisper fallback. No audio leaves the machine.
- Multilingual Whisper with automatic language detection.
- Model deep-dive screen with family wordmarks.
- Voice-UX and TUI polish: opt-in postures, modal screens, refreshed branding (masthead + splash), richer status footer.
- Onboarding + default-model fallback.
Everything runs on the native Swift + MLX engine, no Python dependency.
Install
brew tap srvsngh99/krillm && brew install krillm (or brew upgrade krillm)
Asset: krillm-0.7.0-arm64-apple-macos.tar.gz (Apple Silicon)
Full changelog: see CHANGELOG.md [0.7.0].