Connect an AI model for systemd work

Choose the model that fits your privacy needs. Devnix Ops sends prompts to the provider you pick. Every action appears as a card with its real result.

Choose your model

Cloud AI provider

Use a cloud AI provider through a standard API endpoint. Prompts go to that provider, and file text the agent reads travels with the prompt. You control the endpoint, which suits bigger models for harder steps.

Local PC model

Run a model on your PC. Nothing leaves the machine. Prompts stay local. This suits strict privacy. You need a capable computer. The agent uses the local model. Results appear in the chat. No network is required for the model.

Built-in model

Use the built-in model ready to use without your own endpoint.

Test your connection

The Test button checks an endpoint, shows response time and the context window.

Pick the right model

Use a bigger model for refactors. Use a smaller model for small edits. Send a hard step to a bigger model with run_on_model. The agent decides the path. Changes wait for approval in Ask first mode.

Quality varies between models. Review the diff carefully. In Ask first mode, the agent pauses before actions that change something. In Auto-run mode, actions run immediately. The agent shows the change. You approve the file.