Local AI, one command away

Meet runai

runai picks a model quantization for your hardware, downloads it, and opens a local chat using llama.cpp. No manual GGUF hunting or runtime configuration.

Install on macOS, Linux or WSL

curl -fsSL https://canirun.ai/runai/install.sh | bash

Review the stable installer script before running it.

Run a model

Copy the runai command from any model page, or pass its model ID directly:

runai run qwen3.5-4b