Skip to content

The publication for web craftspeople Tuesday, 22 September 2026

AI for the web

GitHub Copilot adds Grok 4.7 for agentic coding

GitHub has rolled out Grok 4.7, xAI's newest reasoning model, across Copilot Pro, Pro+, Max, Business and Enterprise. Built for agentic coding and multistep workflows, it joins the growing list of models already available across most supported IDEs.

GitHub announced on September 21, 2026 that Grok 4.7, xAI’s latest reasoning model, is now available in GitHub Copilot. The gradual rollout covers the Pro, Pro+, Max, Business and Enterprise tiers, and follows Grok 4.6, which had been available since August.

A model built for long-running tasks

Grok 4.7 joins an already sizeable list of models, at a time when many teams are more focused on stabilizing their command-line tooling than chasing every new release.

According to the announcement published on the GitHub changelog, Grok 4.7 is designed for agentic coding and complex, multistep workflows, building on Grok 4.6. It joins the already substantial list of selectable models in Copilot, alongside models from Anthropic and OpenAI.

Availability across plans and environments

The model is available in Visual Studio Code, Visual Studio, Copilot CLI, the Copilot coding agent, the Copilot app, as well as in JetBrains, Xcode and Eclipse. Because the rollout is gradual, not every eligible account will see it in the model picker right away.

Copilot planAccess to Grok 4.7
FreeNo
ProYes
Pro+Yes
MaxYes
BusinessYes, managed by administrators
EnterpriseYes, managed by administrators

Billing and model policy

Grok 4.7 is billed at the provider’s list pricing under usage-based billing, with no specific rate multiplier detailed in the announcement itself; exact figures should be checked in Copilot’s models and pricing documentation. On Business and Enterprise plans, administrators manage access to the model through the model policy in Copilot settings. By default, new models are enabled automatically unless an administrator has turned off global default enablement or explicitly disabled this particular model.

# List available models from Copilot CLI
copilot models list

# Select Grok 4.7 for the current session
copilot models set grok-4.7

Grok 4.7 is designed for agentic coding and complex, multistep workflows.

On Business and Enterprise plans, automatic enablement of new models means Grok 4.7 can show up for a team without any action from an administrator who isn’t actively monitoring the model policy — along with the usage-based billing that comes with it.

A release cadence that keeps accelerating at xAI

Grok 4.7 succeeds Grok 4.6, available in Copilot since mid-August, and Grok 4.5, released in late July. That roughly monthly cadence between major versions matches what’s already been observed with other model providers integrated into Copilot, and raises the question of tooling stability for teams building production agents around a specific model.

Key takeaways

GitHub once again expands its model catalog with Grok 4.7, available across most paid tiers and every supported Copilot environment. Managing automatic enablement remains the point to watch for organizations that want to keep usage costs under control.

The monthly cadence of new models landing in Copilot is starting to look like a stream that needs active management rather than passive following. The real question isn’t which model tops which benchmark anymore, but whether teams can build an enablement policy that avoids paying for a model nobody actually evaluated before it reached production. — Simon Janvier

Further reading: the official announcement on the GitHub changelog.

Read next