tailoring your AI-powered coding assistant to fit each project’s unique demands just got a little easier with GitHub’s latest feature for Copilot Pro and Pro+ subscribers. The latest update introduces a powerful model picker for the Copilot coding agent. This upgrade puts the power to choose your AI model directly in your hands, offering greater control over how your coding tasks are handled.
Understanding the Model Picker
The model picker appears whenever you start a task with the Copilot coding agent, be it from the agents tab, agents panel, a GitHub issue, or using the Raycast extension. A dropdown menu lets you select from several advanced AI models:
- Auto: Lets Copilot automatically select the best-performing model for speed and reliability.
- Anthropic Claude Opus 4.5: Excels at advanced reasoning and code comprehension.
- Anthropic Claude Sonnet 4.5: Strikes a balance between speed and capability.
- OpenAI GPT-5.1-Codex-Max: The newest OpenAI model, optimized for generating high-quality code.
Choosing "Auto" means Copilot makes smart decisions for you, but manual selection lets you fine-tune performance as needed.
Access and Availability
Currently, the model picker is exclusive to Copilot Pro and Pro+ subscribers. However, GitHub has announced plans to expand access, bringing the feature to Copilot Business and Enterprise customers soon. Look for future integration with more platforms like Visual Studio Code and GitHub Mobile, making it easier to leverage this flexibility across all your development tools.
Why Model Selection Matters
Direct model selection is a game-changer for developers. By choosing the right AI model for each task, you can optimize both speed and output quality. For complex projects requiring deeper reasoning, pick a more advanced model. For everyday coding, a faster model may suffice. This approach supports a variety of workflows and helps teams experiment with the newest AI technologies as they emerge.
Ongoing Improvements in Copilot
GitHub isn’t stopping with the model picker. The Copilot ecosystem is rapidly evolving, with recent updates including public previews of cutting-edge models like GPT-5.1-Codex-Max and Claude Opus 4.5. Expanded support for popular IDEs and mobile platforms means even more developers benefit from these enhancements, reaffirming GitHub’s commitment to innovation in AI-assisted coding.
How to Get Started
If you’re already a Copilot Pro or Pro+ user, accessing the model picker is simple. For detailed, step-by-step instructions, visit the official GitHub documentation on model selection. Stay alert for upcoming updates as the feature becomes available to a broader user base.
Final Thoughts
The model picker for Copilot coding agent signals a new era of customizable AI in software development. By letting developers select their preferred AI model, GitHub empowers teams to optimize their workflow and stay at the forefront of coding innovation.
Source: GitHub Changelog
Unlocking Flexibility: Copilot Coding Agent’s New Model Picker Empowers Developers