GitHub Copilot Now Powered by OpenAI’s Latest GPT-4.1 Model

OpenAI's latest model, GPT-4.1, is now live in GitHub Copilot and GitHub Models, delivering cutting-edge AI capabilities directly to developers' workflows. the company has announced.

GPT-4.1 is rolling out across all Copilot plans, including Copilot Free. Developers can start using it by selecting “GPT-4.1 (Preview)” via the model picker in Visual Studio Code or in GitHub.com chat.

According to GitHub, whether a developer is debugging, modernising legacy code, or writing tests, GPT-4.1 is built to accelerate thier workflow.

Enterprise admins will need to enable GPT-4.1 through a new policy setting in Copilot Admin. Once activated, developers will see GPT-4.1 as an option in the model selector.

Starting later today, users of GitHub Models can access GPT-4.1 in the playground to prototype, refine, and iterate AI-driven applications. It will be available alongside other major models from Cohere, DeepSeek, Meta, and Microsoft, giving developers the freedom to explore and compare capabilities.

Outperforming GPT-4o across the board, GPT-4.1 offers major improvements in code generation, instruction following, and long-context reasoning.

With a refreshed knowledge cutoff of June 2024 and an expanded context window, it’s designed to meet the demands of modern software development.

GPT-4.1 has been fine-tuned using real-world developer feedback to handle practical coding scenarios—improving performance in frontend development, reducing unnecessary edits, and enhancing reliability in formatting, structured outputs, and tool usage.

The result is a faster, more responsive model well-suited for everyday development needs.