Installation

The Quick Start showed the one-liner install. This section covers platform-specific details, system requirements, and building from source. Quira is available for macOS (Apple Silicon and Intel) and Linux. Windows support is planned for Phase 2.

System Requirements

macOS

Quira supports both Apple Silicon (M1/M2/M3/M4) and Intel Macs. The installer auto-detects your architecture.

# One-line install (auto-detects Apple Silicon vs Intel)
curl -fsSL https://get.qu-ira.com | sh

# Or install via Homebrew
brew install --cask quira

On Apple Silicon, Quira uses the MLX runtime for GPU-accelerated local AI. On Intel Macs, it falls back to llama.cpp with CPU inference.

Linux

# Debian / Ubuntu
wget -qO- https://get.qu-ira.com/linux | sudo sh

# Arch Linux (AUR)
yay -S quira-bin

# Fedora / RHEL
sudo dnf install quira

# AppImage (any distro)
wget https://releases.qu-ira.com/latest/quira.AppImage
chmod +x quira.AppImage
./quira.AppImage

On Linux, Quira uses llama.cpp with CUDA (NVIDIA), Vulkan, or CPU fallback for local AI inference. For GPU acceleration, ensure your CUDA or Vulkan drivers are up to date.

Build from source

Building from source requires Rust (1.75+), Node.js (20+), and the just command runner.

git clone https://github.com/tennnennsui/quira.git
cd cortex
just setup    # Install dependencies
just build    # Build release binary
just run      # Run the development build