Connect Ollama

Install Ollama from https://ollama.com/ and download a model suitable for your computer. Start Ollama, then enable Local AI in Plugins and open AI provider settings. Select the local connection, confirm the usual endpoint http://localhost:11434 and choose an installed model. Test the connection before asking about your writing.

Model downloads can be large and inference depends on memory/processor capacity. If connection fails, confirm Ollama is running and the model exists; do not paste a cloud key into a local model name. A remote Ollama server is not on-device processing. Generative use requires Pro even when the model service is local.