// Getting Started

Installation

Deploy Uneven AI globally on your machine using standalone, hardware-accelerated executables. No Node.js runtime required.

Universal One-Line Installation

Our automated network bootstrap identifies your operating system and CPU architecture, downloads the latest optimized opaque binary over TLS, and sets up your environment variables transparently.

Linux & macOS

Run the following pipeline in any standard Unix shell terminal:

bash
curl -fsSL https://uneven.dev/install.sh | bash

Windows 11

Open PowerShell as Administrator or Standard User and execute:

powershell
irm https://uneven.dev/install.ps1 | iex

What the Installer Does

To guarantee complete auditability and continuous security integration, our open bootstrap script performs the following atomic operations locally:

  • Creates the global binary base directory at ~/.uneven/bin
  • Resolves the corresponding platform runtime core (linux-x64, darwin-arm64, or windows-x64)
  • Fetches the payload directly from our secure CDN proxy edge infrastructure
  • Appends the binary path to your active profile configuration (.bashrc, .zshrc, or Windows PATH registry)
  • Verifies cryptographic hash consistency automatically before granting execution permissions

Manual Binary Verification

Once installed, verify successful setup and hardware acceleration status by printing our baseline runtime info:

terminal
uneven-ai --version

If your terminal session cannot find the executable immediately, refresh your current shell profile environment:

terminal
# Bash
source ~/.bashrc

# Zsh
source ~/.zshrc

Updating Your Installation

Uneven AI supports native zero-downtime atomic self-updates. Simply run the built-in update tool from anywhere:

terminal
uneven-ai update
Quick Start →