Skip to main content
Use Ollama to serve models locally or from another computer on your network. The model server must be running and have a compatible model available.

​
Connect on the same Mac

  1. Start Ollama and download a model using its own model-management tools.
  2. Open Settings → AI Model → Ollama.
  3. Configure its credential/server URL. The default preset uses http://127.0.0.1:11434.
  4. Select a model from the available list and try a short text-only request.
  5. Select that model in an Agent or conversation when you want to use it.

​
Connect to another computer

Use the server computer’s LAN address instead of 127.0.0.1 or localhost; those names always refer to the computer running Enconvo. Configure the model server to accept connections on its local-network interface, and enable Enconvo under System Settings → Privacy & Security → Local Network if macOS blocks access. Keep the port and API path appropriate to your server. Allow access only on networks and interfaces you intend to use.

​
Troubleshooting

Local model inference does not make remote tools offline. See Local Models for the full processing chain.