Ready to deploy a high-performance LLM infrastructure on your own hardware? Choose your platform below.
AegisLLMServer is a comprehensive package including multiple LLM backends and pre-optimized libraries. The download size is approximately 4.5GB. Depending on your network speed and hardware, installation may take several minutes. Please ensure you have at least 10GB of free disk space before proceeding.
Download the Debian package and use the terminal to install it. This ensures all dependencies are correctly handled.
Download1. Open your terminal.
2. Navigate to your Downloads folder:
3. Install the package using apt (recommended):
Note: Using sudo apt install will automatically download any required system
dependencies.
The server will automatically detect your GPU (CUDA) or fall back to high-performance CPU inference if no compatible GPU is found.
Native Windows installer is currently in development.
We are currently working on a native Windows installer with full DirectML support for NVIDIA and AMD GPUs.
Optimized Metal-based backend for M1/M2/M3 chips is currently in closed beta testing.