Connect an AI Model

Connect DevSpan to the AI model you choose.

Choose your model source

Cloud AI provider

Use a cloud AI provider through a standard API endpoint. Your prompts go only to that provider, so file text leaves your PC. You must manage your own API keys and quotas. The agent sends requests over the network to complete tasks.

Local PC model

Run a model locally on your PC. With a local model nothing leaves the machine; you manage the model software.

Built-in model

The built-in model is ready to use without your own endpoint.

Test your connection

Press the Test button to check an endpoint. The tool shows response time and the context window.

Pick the right model size

Send a hard step to a bigger model using run_on_model. Use a larger model for complex refactors or deep logic analysis. Smaller models work fine for small edits. The agent chooses the tool based on your task. You control which model handles which step.

Quality varies between different model providers. Always review the diff before accepting changes. Use Ask first to pause before every action that changes something. You see the exact file changes. You decide if the output is correct.