Developers who use Copilot now have another model to try. GitHub has started rolling out Grok 4.7, from xAI, and presents it as an option for agentic coding and complex tasks with several steps.
Think of a change that starts by reading files, moves through edits in different parts of a project, and ends with tests and fixes. That is the kind of work where trying the new model makes sense. The announcement describes what Grok 4.7 is designed for, but it does not show how it compares with other models in your repository.
Where to try it
Grok 4.7 is coming to the model picker in VS Code, Visual Studio, JetBrains, Xcode, Eclipse, Copilot CLI, Copilot cloud agent, and the GitHub Copilot app. Eligible plans are Pro, Pro+, Max, Business, and Enterprise.
The rollout is gradual. If you do not see it yet, check again later. If you work across several tools, confirm availability in each one before planning a test that starts in the editor and continues in the CLI or cloud agent.
A useful first test
Give the new model a task your team already understands: fix a bug with tests, update an integration, or make a small refactor across several files. Then run a comparable task with the model you use today.
Look at whether the code passes tests, how much review or rework it needs, and how many attempts it takes to get there. In agentic workflows, a convincing answer in chat is only the start. What matters is how the model works through the project and responds to errors.
Cost belongs in that comparison too. GitHub says Grok 4.7 uses usage-based billing at provider list pricing. Its models and pricing documentation has the rates and plan conditions. Compare the consumption of the whole task, especially when the model makes several rounds of reading, editing, and testing.
For Copilot Business or Enterprise administrators, model policy is another detail to check. New models may be enabled by default unless the global setting or a specific model setting blocks them. Check the actual policy before making the model available to the team.
This release gives developers another choice inside tools they already use. A small test on your own code can answer the adoption question: does Grok 4.7 complete the task with less rework, at a cost that makes sense?