Requirements
Requirements
| Item | Requirement or Recommendation |
|---|---|
| Memory (RAM) | The script prints how much memory your machine has and shows the download size of each option next to it. As a rough rule, the model should be comfortably smaller than your RAM. Choosing a model that’s too big won’t break anything, but it will be painfully slow. Besides the model itself, RAM is also needed for the cache while it runs |
| Operating System | Any (macOS, Linux, or Windows). The script auto-detects your OS. On macOS it can install llama.cpp via Homebrew or the official direct installer; on Linux and Windows it uses the official direct installer. No Homebrew required. |
| Disk Space | Recommended Between 5 GB and 20 GB, depending on which model you pick. |
| Patience for one step | The model download is several gigabytes. It can take anywhere from a few minutes to an hour depending on the quality of your connection. |
| Internet Connection | Needed only during setup. Afterwards, the assistant works offline. |
You can also choose skip at the model step and install nothing yet — steps 3 and 4 still run, and llama serve downloads a model on its first run.