If you write code, or you work near people who do, you may have noticed that AI coding helpers keep multiplying. This update from GitHub is worth a look because it changes where you can use Copilot Fusion and which AI models sit behind it.
What Copilot Fusion is
GitHub Copilot is an AI assistant for writing and reviewing software. Copilot Fusion is one of the ways you can use it. Instead of you picking a single model, Fusion lets Copilot choose and coordinate several AI models for each task. Think of it like a team lead who decides which specialist should handle each piece of the job.
You still have control over how hard the system works. According to GitHub, you set an effort level, and Copilot handles the rest.
What's new in this update
GitHub's announcement covers three main changes:
- New models added to each tier. Claude Sonnet 5.5 joins Copilot Fusion Max, GPT-6.1 Terra joins Copilot Fusion High, and MAI-Code-2 joins Copilot Fusion Standard.
- More places to use it. Fusion is now available on the latest versions of github.com, in Visual Studio, and in the GitHub Mobile beta apps for iOS and Android. GitHub says support in the production GitHub Mobile apps is coming shortly.
- Public benchmark listings. Copilot Fusion now appears on Terminal-Bench 4.1 leaderboards hosted by Stanford, Harbor, and the Laude Institute.
What Terminal-Bench is, and why the leaderboard matters
A benchmark is a standard test that everyone runs the same way. Terminal-Bench 4.1 checks how well AI agents (AI programs that can take actions, not just answer questions) complete hard, realistic tasks inside a terminal, which is the text-based command window developers use to run their projects.
GitHub reports that Copilot Fusion Max reached a 71.2% resolution rate on this test. That number is GitHub's own claim, so treat it as the company's result. The more useful part for most readers is that independent leaderboards let anyone see how Copilot performs next to other agents on the same tasks, under the same rules.
Wrap-up
GitHub's update makes Copilot Fusion available in more places and adds three new models to its tiers. The benchmark listings give you a public way to check its performance. Your next step is simple: visit GitHub's official Copilot documentation today and see whether Fusion fits the way you work.
