GitHub presents a demo by @shanselman illustrating how developers can switch between multiple AI models using the Copilot CLI, highlighting features such as model choice, reasoning effort, and voice dictation.

Switching Models in GitHub Copilot CLI: Demo by @shanselman

In this video demo, @shanselman demonstrates the advanced capabilities of the GitHub Copilot CLI. The focus is on the power of choice that developers have when working with Copilot CLI:

Switching Between Models: The CLI supports 14 different models, including not only advanced OpenAI models like GPT 5.2 Codex, but also Claude Opus 4.5 and others. This allows developers to select the model best suited for their coding style or specific task.
Real-World Scenario: The demo covers planning and executing a complex upgrade for a Next.js application. It illustrates how model switching can impact the reasoning and results provided by the Copilot CLI.
Adjusting Reasoning Effort: Users can adjust how much reasoning or analysis the chosen model applies when generating responses, allowing for tailored output depending on complexity and need.
Voice Dictation Support: Copilot CLI includes voice dictation functionality, making the prompting process faster and more natural, especially during hands-on development sessions.

Key Features Shown

Switching among 14 supported models
Integrating Copilot CLI with developer workflows for application upgrades
Tuning model effort for varied complexity
Utilizing voice prompts for efficiency

For further resources and updates, connect with GitHub through their YouTube, Blog, X, LinkedIn, Newsletter, Instagram, and TikTok.

About GitHub: Over 180 million developers use GitHub to create, share, and ship software worldwide. For more information, visit https://github.com.