GitHub adds Grok 4.5 to Copilot for agentic coding

GitHub is rolling out xAI’s Grok 4.5 reasoning model in GitHub Copilot. The model is aimed at agentic coding and complex multi-step work, with support for text and image input, a context window of up to 500,000 tokens, and low, medium and high reasoning-effort settings.

The model will appear in Copilot Pro, Pro+, Max, Business and Enterprise plans. GitHub says users will be able to choose it in Visual Studio Code, Visual Studio, Copilot CLI, Copilot cloud agent, the GitHub Copilot app, JetBrains, Xcode and Eclipse. Business and Enterprise administrators need to enable the Grok 4.5 policy before teams can use it.

GitHub says its internal testing found Grok 4.5 especially useful for terminal-based coding tasks, parallel tool use and direct action in VS Code and Copilot CLI. The model is billed at provider list pricing under usage-based billing, so the practical question for teams is not just whether it performs well, but when it is worth the extra model cost.

The takeaway is that coding assistants are becoming model routers. Developers are no longer choosing one assistant as much as choosing a workspace where several frontier models compete on long-context code understanding, tool use, price and reliability.