UpdateImportance 3 of 5
GitHub Copilot Fusion expands with new models, wider availability, and Terminal-Bench results
GitHub announced updates to GitHub Copilot Fusion, including new models, availability on more Copilot surfaces, and results on two public Terminal-Bench leaderboards. The source post is a short excerpt, so specific model names and benchmark scores are not confirmed here.

Why it matters
Builders using GitHub Copilot may get access to additional models and surfaces for Fusion, and public Terminal-Bench results offer a reference point for comparing coding-agent performance.
Who should care
Software developers and engineering teams using GitHub Copilot, and teams evaluating AI coding tools.
What you can do
Open the GitHub changelog post, check which Copilot surfaces now include Fusion, and compare the published Terminal-Bench results against your current coding tool.