From install to offline chat in five minutes

No code, no terminal, no account. Here is the full path from installing an app to chatting with a real LLM that runs in airplane mode, step by step.

Step one: install the runner

You need an app that can load and run model files, and Local AI is built exactly for that job. Install it, open it, and there is no sign-up wall - the app works fully without an account of any kind. The only real requirement is free storage: leave yourself a few gigabytes so model files have room to land on your device.

Step two: pick a model

Open the model browser inside the app and choose something to download. For a first try, pick a model between 1B and 3B parameters - big enough to write well, small enough to run on most phones. Tap download, wait for it over Wi-Fi, and the model is yours permanently. Bigger models up to 12B are sitting there whenever you want them.

Step three: chat, then go dark

Start a conversation like any chat app - the answers stream in word by word as the model writes. Then comes the moment that proves it works: flip on airplane mode and ask another question. If it still answers, the model is genuinely running on your device, and that is the whole setup. No server, no subscription, nothing else to configure.

Worth knowing before you dive in

Two tips smooth the experience. Match the model size to your phone - if replies crawl, drop down a size rather than suffering through slow answers. And explore past plain chat: vision models can look at your photos, and a built-in voice assistant answers spoken questions. All of it works offline, so every feature survives the next dead zone you hit.