Connect your AI model

You choose between a cloud AI provider, a local model, or the built-in model.

Choose your model source

Cloud AI provider

Use a standard API endpoint to connect to a cloud AI provider. Your prompts travel over the network to that provider. File text and context leave your machine during the request. You control the endpoint and the provider may charge separately for its use.

Local PC model

Run a model directly on your own computer hardware. No network traffic leaves your machine for the prompt data when using a local model. Your SQL exercises stay completely private and local.

Built-in model

Start immediately without configuring any external endpoint. The built-in model is ready to use without your own endpoint.

Test your connection

Press the Test button to verify your endpoint works. The agent checks the endpoint and shows response time and context window size. This confirms the model is reachable and ready for your SQL queries.

Pick the right model for the task

Send a hard step to a bigger model with run_on_model. Use smaller models for simple syntax edits or quick checks.

Quality varies between models and your specific dataset. Always review the diff before accepting changes. Use Ask first to pause before any data-changing actions. You control the final decision on every edit or query.