AI providers

Roleplay Manager does not include a model. It connects to the provider you choose and sends the prompt to it.

Ollama (local)

1. Install Ollama and pull a model, for example ollama pull llama3.1.
2. In Roleplay Manager, open System → Providers.
3. Add http://localhost:11434 as an Ollama provider.
4. Use Test connection, save and choose the default model.

Everything stays on your computer. Performance depends on your hardware; a smaller model is a good starting point if you do not have a dedicated GPU.

OpenAI-compatible endpoint

1. Open System → Providers.
2. Add an OpenAI-compatible provider with its base URL and API key.
3. Test the connection and choose a model.

This works with LM Studio, OpenRouter, a self-hosted gateway or any other service that implements the OpenAI chat API.

Note

If you use a remote provider, only the prompt you send leaves your machine. The provider’s terms and privacy policy apply to that request.

Models per conversation

The default model is used for new chats, but each conversation can use a different provider and model. Open a conversation’s settings to change it without affecting the rest.

Troubleshooting

• Connection refused with Ollama: make sure Ollama is running and the URL is http://localhost:11434.
• 401 or 403: check the API key and the provider’s base URL.
• The model is not listed: refresh the model list after pulling or enabling it in the provider.