Grok 4.7 rolls into GitHub Copilot with usage-based billing

GitHub is adding xAI's Grok 4.7 to Copilot, extending the list of reasoning models developers can choose inside their coding tools. The rollout announced on September 21 is gradual, and the model will be billed at the provider's list pricing under usage-based billing. A developer who cannot yet see it in the picker may simply be waiting for access to arrive.

The practical story is less about a new chatbot and more about where a model can be used. GitHub says Grok 4.7 is available across editors, its command-line tool and the cloud agent. It describes the model as designed for agentic coding and complex, multistep work, but the announcement gives no independent benchmark or measured comparison with other Copilot choices.

One model across several work surfaces

The eligible subscriptions are Copilot Pro, Pro+, Max, Business and Enterprise. GitHub lists Visual Studio Code, Visual Studio, JetBrains, Xcode and Eclipse among the places where users can select the model. It also names Copilot CLI, the Copilot app and the cloud agent.

That spread matters for teams whose workflow moves between a local editor and an asynchronous coding agent. A consistent model choice may simplify experiments across those surfaces. It does not guarantee identical results: task context, tool access and workflow differ between an editor interaction and an agent working in the cloud.

GitHub's model picker makes the addition visible to users, while the release note says it may not appear for everyone at once. Developers should check the actual available models on their own plan and surface. The note does not state a universal completion date for rollout.

The price and control details

Usage-based billing at provider list pricing is an important distinction from assuming every model is covered in the same way by a flat subscription. GitHub points users to its current models and pricing documentation for the charge details. A team comparing models should include cost as well as task success, latency and the amount of human correction required. The release note itself does not supply enough evidence to rank Grok 4.7 on those measures.

Business and Enterprise administrators can manage access in Copilot's model policy settings. Under GitHub's default model enablement, a new model is automatically enabled unless the administrator has disabled the global default or explicitly blocked that model. Organizations with approval or data-governance requirements should inspect their setting before treating the rollout as an individual user choice.

This is also a reminder that a model appearing in a familiar interface does not remove the need for code review. Suggested patches can be useful starting points, while tests, dependency checks and a developer's understanding of the change remain the basis for trusting it.

What to watch after rollout

GitHub's brief announcement establishes availability, surfaces, plan eligibility and billing. It does not establish whether Grok 4.7 writes better code than any alternative or how it performs on a particular repository. Those questions require repeatable tasks and direct comparison under the same conditions.

For now, the change gives Copilot users another model to test where they already work. The test worth running is concrete: on a real multistep issue, does the new option reduce review and rework enough to justify its usage cost?