Auto Model Selection Now Live

GitHub Copilot's auto model selection feature is now generally available in JetBrains IDEs for all Copilot plans. This intelligent routing system automatically selects the best model on your behalf based on real-time availability and performance metrics, eliminating the need for manual model selection while maintaining reliability even during peak usage periods.

How Auto Selection Works

The auto feature dynamically routes requests between models including GPT-5.4, GPT-5.3-Codex, Sonnet 4.6, and Haiku 4.5 (depending on your plan and organization policies). Key capabilities include:

• Transparency: Hover over model responses to see which specific model was used
• User control: Switch between auto and manually selected models at any time
• Policy compliance: Respects all administrator-configured model restrictions

Billing and Discounts

Premium requests using auto are billed based on the selected model's multiplier rate. All paid subscribers receive a 10% discount on the model multiplier when using auto—for example, models with a 1x multiplier will consume 0.9 premium requests instead of 1.

Future Enhancements

GitHub plans to make auto even more intelligent by enabling task-aware model selection. The system will soon match model capabilities to request complexity, automatically routing simple queries to efficient models while assigning complex tasks to more capable ones.

Developers can share feedback in the Community Copilot Conversations discussion.