Grok 4.7 in GitHub Copilot is rolling out, GitHub announced on September 21, 2026. GitHub describes xAI’s new reasoning model as designed for agentic coding and complex, multistep work.
What changed
The official GitHub announcement lists availability for Copilot Pro, Pro+, Max, Business, and Enterprise. Supported model-picker destinations include VS Code, Visual Studio, Copilot CLI, the cloud agent, the Copilot app, JetBrains, Xcode, and Eclipse. Availability is gradual, so the option may not appear immediately.
Business and Enterprise administrators can control access through Copilot model policies. GitHub says new models are enabled automatically under default model enablement unless an administrator disables the global default or the individual model. Usage-based billing follows provider list pricing.
Why this matters for QA engineers
QATechTools analysis: A new model option is a reason to rerun your coding-assistant evaluation. It is not evidence that generated tests will detect more defects. Compare a small, fixed set of known bugs using the same repository revision, prompts, and test environment.
For each proposed fix, check that a regression test fails before the patch and passes afterward. Review assertions for meaningful coverage, record the selected model, and compare execution time and usage cost. Keep human review and CI checks in place before adopting the model for broader automation work.
Verified September 21, 2026. Source: GitHub Changelog, September 21, 2026. QA evaluation guidance is editorial analysis.