Chat with local AI from your phone
LM Mini connects your phone to Ollama and LM Studio at home — or runs models on-device. Private local AI for iPhone, Apple Watch, Android, and Mac.
LM Mini is a private AI chat app for iPhone, Android, and Mac, with an Apple Watch companion. It runs models on your device, connects to LM Studio, Ollama, JAN AI or Unsloth on your Wi‑Fi, and with Pro uses LM Mini Home → Share with phone on Mac so your desk GPU is reachable from anywhere.
Pick the path that matches how you already run AI — then get the app.
Chat offline with models on your phone or Mac. No PC required. Free includes small starter models; bigger and MLX models need Pro.
Browse models → Find your phone →Use the big model on your home GPU through LM Studio or Ollama — with a phone UI that does not fight you.
Wi‑Fi setup →With Pro, pair once with LM Mini Home → Share with phone on Mac. Encrypted relay — no port forwarding or VPN. Windows and Linux use Connect.
Remote / Connect →Local AI first — Personas, group chat, voice, tools, and image gen when you need them.
Open source
The free app's source is on GitHub under the MIT License: chat, on-device models, LM Studio and Ollama, personas and voice. Read what it does, build it yourself, or send a fix. Pro features stay in the official App Store, Google Play and Mac builds, and they fund development.
Away from home
With LM Mini Pro: on Mac, download LM Mini Home, open Share with phone, scan once. Windows and Linux use LM Mini Connect. Same-Wi‑Fi LM Studio and Ollama never needed a companion app.
Plain-language notes from Sultan Haider — product, local AI, and why we ship the way we do.
Short answers before you install — more detail lives in the docs.
Yes. Download free on the App Store and Google Play. Pro is optional for remote access, Group Chat, bigger on-device models, cloud providers, App Lock, and some premium tools.
Yes, the core is. The free app is MIT-licensed on GitHub. Pro features ship only in the official store builds. How open core works.
Not to chat once the model is downloaded. Same for Ollama. You need the internet to download models, and a path to your PC if you want that model on your phone away from home. Full post.
No. You can chat fully on-device. LM Studio and Ollama are for when you want bigger models on a home PC.
Local and on-device chats stay on your devices. Remote Share with phone (Pro) uses an encryption key from QR pairing that only you control.
Deprecated — it keeps working through the end of 2026. Use LM Mini Home. Windows and Linux Connect stay on the Connect page.
LM Mini
iPhone, iPad, Apple Watch, Android, and Mac — on-device or connected to the PC you already trust.
