Kivarro v0.1.0-alpha

Kivarro local model inference workstation alpha release.

Downloads:

• Windows x64: use the Windows setup executable or MSI.
• Windows ARM64: use the ARM64 setup executable or MSI.
• macOS: use the DMG for your architecture.
• Linux: use the AppImage, Debian package, or RPM produced by the Linux build.

Notes:

• This is an unsigned alpha build. Windows SmartScreen and macOS Gatekeeper may warn on first launch.
• Kivarro supervises local backends such as llama.cpp llama-server; install a backend separately and point Kivarro at it if it is not on PATH.
• Do not attach private prompts, documents, model files, or unsanitized logs to public issues.