Set up Ollama

MindMap Chat talks to a local AI server called Ollama. Run it on your own machine, or on one capable PC that serves every device on your network. Either way, your map content never has to touch the public internet.

On desktop the app can install and start Ollama for you — press Save in Settings → AI and it downloads Ollama, launches the service, and pulls your chosen model. If you'd rather set it up by hand, or you want to share one server, read on.

Install Ollama yourself

macOS & Windows

Download the installer from ollama.com/download and run it. Ollama installs a background service that listens on http://localhost:11434.

Linux

curl -fsSL https://ollama.com/install.sh | sh

Pull a model

From a terminal, download a model to run:

ollama pull llama3.1
# or a smaller, faster option:
ollama pull qwen2.5

Point the app at it

Open Settings → AI, keep the Base URL as http://localhost:11434/v1, pick the matching Model, and press Save. The model's badge should read installed.

One server for your whole network

You don't need Ollama on every device. Run it once on a single capable PC — plenty of RAM, and ideally a GPU — and let every other laptop, tablet, and phone on the same network use it. This is the easiest way to get good local AI on machines that couldn't run a model themselves.

On the host machine

Tell Ollama to accept connections from other devices by setting OLLAMA_HOST=0.0.0.0 in its environment, then restart the Ollama service. By default Ollama only listens on localhost, which other machines can't reach.
- Linux (systemd): run systemctl edit ollama.service and add Environment="OLLAMA_HOST=0.0.0.0", then systemctl restart ollama.
- macOS: launchctl setenv OLLAMA_HOST 0.0.0.0 and restart the Ollama app.
- Windows: add a user environment variable OLLAMA_HOST = 0.0.0.0, then restart Ollama.
Pull the models you want to share (see above).
Allow inbound TCP port 11434 through the machine's firewall.
Note the host's LAN address, for example 192.168.1.50.

On every client device

Open Settings → AI.
Set the Base URL to http://192.168.1.50:11434/v1 (your host's address).
Pick a model the host has pulled, and press Save.

Clients don't install anything — they just point at the shared host. The app shows a small notice under the Base URL confirming the endpoint is on your network.

Keep your endpoint on your own network

Where Ollama runs decides where your data goes. Everything the AI sees — your map outline, notes, and chat — is sent to whatever Base URL you configure. MindMap Chat tells you which of three situations you're in:

On this device (localhost / 127.0.0.1) — nothing leaves the machine. The most private option.
On your LAN (a 192.168.x, 10.x, or *.local host) — content stays on your own network. The app shows an info notice; use only hosts you trust, since anyone who can reach the server can read your prompts.
On the public internet (a routable IP or a domain like ollama.example.com) — content leaves your network entirely, and the app shows a strong warning.

⚠ Don't point the app at an Ollama endpoint on the public internet unless you run and trust that server.

Doing so streams your private map content off your network, where it may be logged or retained. The same caution applies to any remote, API-key-protected endpoint (such as a hosted OpenAI-compatible service): a remote endpoint is off-network whether or not it needs a key. Your local database is always encrypted at rest, but that protection does not extend to content you send to a remote endpoint — which is exactly why the endpoint you choose matters.

When sharing over a LAN, prefer plain HTTP only on a network you control, and treat anything reachable from outside your network as public.

Troubleshooting

Clients can't connect to the shared host. Almost always either OLLAMA_HOST=0.0.0.0 isn't set (Ollama is still bound to localhost), or port 11434 is blocked by the host's firewall.
404 / model not found. The client asked for a model the host hasn't pulled. Run ollama pull <model> on the host.
Timeouts. The host is unreachable — check the IP address and that both devices are on the same network.

More in the user guide. Free, no account — download the app.