LM Studio and cloud providers

In LM Studio, load a compatible model and start its local OpenAI-compatible server. In The Novelist’s provider settings choose LM Studio, verify the endpoint (normally http://localhost:1234), and use the model identifier offered by the server. The local preset does not normally require an API key.

Cloud presets include OpenAI, DeepSeek, OpenRouter, Groq and Mistral, plus a custom compatible endpoint. Use your own provider account/key, valid base URL and supported model; billing is separate. Selected prompt/context leaves your computer. Check the provider’s privacy/retention terms and do not share API keys in support screenshots. Connection success does not guarantee every model supports the requested feature.