GitHub is rolling out xAI’s Grok 4.5 in GitHub Copilot, giving developers another frontier coding model inside the Copilot model picker.
The July 28 changelog says Grok 4.5 will be available to Copilot Pro, Pro+, Max, Business, and Enterprise users. GitHub lists Visual Studio Code, Visual Studio, Copilot CLI, Copilot cloud agent, the Copilot app, JetBrains, Xcode, and Eclipse among the surfaces where users will be able to select it.
This is not a new Grok 4.5 launch. The AI Feed covered SpaceXAI’s July 8 model release earlier this month. The new piece is distribution through Copilot, plus the enterprise controls around that distribution.
Copilot is becoming the model marketplace for coding work
GitHub describes Grok 4.5 as a reasoning model for fast, agentic coding and complex multi-step workflows. The changelog says it supports text and image inputs, low, medium, and high reasoning effort, and a context window of up to 500,000 tokens.
The rollout also follows GitHub’s July 27 governance updates for the Copilot app and cloud agent. That timing matters. As Copilot adds more models, the admin layer becomes the place where organizations decide which model can touch which work.
For Business and Enterprise plans, GitHub says administrators must enable the Grok 4.5 policy in Copilot settings, and that the policy is off by default. The model is billed at provider list pricing under usage-based billing.
Model choice now has governance cost
For individual developers, the product change is simple: another model appears in the picker as rollout reaches the account.
For organizations, the choice is less casual. Model access affects cost accounting, security review, data-handling posture, and reproducibility of agent work. A task completed by Grok 4.5 may have a different price, reasoning behavior, context limit, and review profile than the same task run on Claude, Gemini, or OpenAI models inside Copilot.
The practical result is that model selection is becoming a managed policy decision. Enterprise teams should expect to document why a model is enabled, who can use it, and which work should stay on approved defaults.
