___ ___                             .____    .__  __
 /   |   \ ___.__.______   ___________|    |   |__|/  |_  ____
/    ~    <   |  |\____ \_/ __ \_  __ \    |   |  \   __\/ __ \
\    Y    /\___  ||  |_> >  ___/|  | \/    |___|  ||  | \  ___/
 \___|_  / / ____||   __/ \___  >__|  |_______ \__||__|  \___  >
       \/  \/     |__|        \/              \/             \/

@hyperlite-ai/darwin-arm64

Native macOS Apple Silicon binary for HyperLite — terminal-native local AI, offline and Metal-accelerated.

This is a platform binary package. Install the main package:

npm install -g hyperlite-ai
hl

→ hyperlite.org

About HyperLite

HyperLite is a local AI assistant built in Rust that runs entirely on your hardware. No cloud, no API keys, no telemetry. On Apple Silicon, inference runs through a Metal-accelerated llama-server installed via Homebrew — using the full unified memory bandwidth of M-series chips for fast local LLM inference.

This package

Contains a single precompiled binary: hl

• Format: Mach-O ARM64, macOS
• Target: aarch64-apple-darwin
• Compiler: Rust stable, built on macOS 14 (Sonoma)

Requirements

• macOS 11+ on Apple Silicon (M1 / M2 / M3 / M4)
• Node.js 16+ (launcher only)
• Homebrew recommended (for automatic runtime install)

Usage

npm install -g hyperlite-ai
hl

On first launch, HyperLite installs llama.cpp via Homebrew for Metal-accelerated inference and downloads a model sized to your RAM.

Gatekeeper note

macOS may block unsigned binaries on first run. If you see a security warning:

xattr -d com.apple.quarantine $(which hl)

Or go to System Settings → Privacy & Security → Open Anyway.

→ hyperlite.org

License

MIT