LM Studio remote access from your phone (same Wi‑Fi or away from home)

In the last post we talked about on-device AI, where the phone is the whole computer. This one is the other half. Your PC already runs LM Studio or Ollama. Your phone is how you talk to it.

People search this a few different ways: LM Studio remote access, connect to a remote model, LM Studio server on the local network, connect to a remote server. Same job. You want the model on your desk, and a decent chat app in your hand.

Two setups, depending on where you are.

Setup 1: same Wi‑Fi (the local network)

This is the fastest path. Phone and PC on the same network. LM Mini talks to your machine directly. You are not putting LM Studio on the public internet. You are letting it serve on your local network, which is what that setting is for.

What you need

• LM Mini on your iPhone, iPad, or Android
• LM Studio or Ollama running on your computer
• Both devices on the same Wi‑Fi

Steps

1. On your PC, start LM Studio or Ollama and load a model.
2. In LM Studio, turn the local server on, and turn on Serve on Local Network. If you skip that second toggle, the phone will not see it. Ollama should already be listening on its usual port.
3. Find your computer's local IP (something like 192.168.1.100). Windows: ipconfig in Command Prompt. Mac: System Settings → Wi‑Fi → Details, or ipconfig getifaddr en0 in Terminal.
4. Open LM Mini, go to Settings, then the server / provider section.
5. Enter the address and port. LM Studio is usually http://YOUR-IP:1234. Ollama is often http://YOUR-IP:11434 if you did not change it.
6. Save, tap Test if the app offers it, start a chat.

If it connects, you are done. Same models, same privacy story, better screen in your hand. This is LM Studio on the local network, not a remote server on the open internet.

When same Wi‑Fi is enough

• You are at home, or anywhere both devices share a network.
• You do not need access from cellular or a coffee shop.
• You want the simplest setup, no subscription.

Setup 2: away from home (actual remote access)

This is the one people mean by LM Studio remote access when they are not on the house Wi‑Fi. The model stays on your desk. You leave. You still want that model, not a tiny on-device one.

On Mac, do not install Connect for new setups. Install LM Mini Home, open Share with phone, scan from the phone. Same QR ritual. Same encryption. Mac Connect keeps working through the end of 2026 (why).

On Windows / Linux, LM Mini Connect is still the companion.

You are connecting to your own remote server: your PC. Not a random host in a datacenter.

What you need

• Everything from Setup 1, plus LM Mini Home (Share with phone) or Connect on Windows/Linux
• LM Mini Pro on the phone for remote access (on-device and same-Wi‑Fi still work without Pro)

Steps (Mac)

1. Open Share with phone on the Mac that runs LM Studio, Ollama, or on-device models.
2. Show the QR. That code includes a session ID and an encryption key you control.
3. On your phone: LM Mini → Settings → Remote Access → Scan QR Code.
4. Scan. Wait for connected. Send a test message from cellular if you want proof it is really remote.

Windows/Linux: same scan flow, QR comes from Connect. Point it at localhost:1234 for LM Studio, add your API token if you use one.

That is the whole ritual. No opening ports on your router. No VPN config with six screenshots and a prayer.

Encryption in plain English

I covered this in the birth story post, but people still ask.

When you pair with that QR code, traffic between your phone and your PC is encrypted with a key that lives with you. The relay forwards encrypted packets. We are not sitting there reading your chat as plain text. Remote access is Pro because running that relay is not free. The privacy model is the same either way.

LM Studio or Ollama?

Both work. Use whichever desktop stack you already have.

• LM Studio: load a model, start the server, Serve on Local Network, connect as above.
• Ollama: run your models in Ollama, point LM Mini at the host and port on Wi‑Fi, or go through Share with phone / Connect for remote.

Same app. Pick the stack you already installed.

Quick troubleshooting

• Cannot connect on Wi‑Fi: Same network? Firewall? Is Serve on Local Network actually on?
• Wrong IP: Laptops switch Wi‑Fi. Re-check.
• Scan works, chat fails: Is LM Studio reachable from the Mac or Connect host? Try a fresh QR.
• It worked yesterday: PC asleep, or Share with phone not running? Re-scan.
• Pro confusion: Local Wi‑Fi does not need Pro. Remote via Home or Connect does.

USB mode bonus (iOS, free)

If you are on iPhone plus Mac and Wi‑Fi is terrible, some builds support USB Mode. App Store Mac may require Wi‑Fi/QR instead. Niche, useful when it works.

Which setup should you use?

• At home on the local network: Setup 1. Fast, free, boring in a good way.
• Out in the world, want your home GPU: Setup 2 plus Pro.
• No PC: stay on-device, from the earlier post.

I use all three depending on the day. That is kind of the point.

If you were searching how to give LM Studio internet access, that is a different question — chat offline vs downloads vs remote phone access are not the same job. For picking what actually fits in your pocket, see which models run well on your phone.

Get LM Mini: lmmini.com