This article is a technical explanation and implementation example created using AI. Although the code and procedures presented are based on primary sources, they have not been verified on actual hardware by the author. Behavior may vary depending on the environment and version. Based on the official changelog, we organize the positioning, available plans, supported interfaces, and administrator configuration methods of the newly added OpenAI GPT-6 model family (GPT-6 Sol and GPT-6 Luna) in GitHub Copilot so that they can be understood safely.
Introduction and Background
In GitHub Copilot, the model family is being expanded so that users can select the optimal model for their development tasks. In addition to the already provided "GPT-6 Astra," "GPT-6 Sol" and "GPT-6 Luna" have now been newly added.
When using generative models, it is important to select the appropriate model according to the purpose and cost. The primary information explains the characteristics of these added models and which plans and environments they can be used in. This article examines the official documentation and reviews the characteristics of each model and the prerequisites for use.
Prerequisites and Notes
All contents explained in this article are official information based on the GitHub Changelog dated September 22, 2026.
Since actual hardware verification has not been performed, results are not shown. Specific behavior in Windows environments and similar setups is marked as [Planned for verification in Windows environment] and [Before actual hardware verification].
Because the rollout is performed gradually, the models may not appear immediately depending on your environment.
There are limitations on billing models and available plans.
Components of the OpenAI GPT-6 Family
According to the primary source, two new models have joined the existing GPT-6 Astra in the GPT-6 family. Their respective characteristics are as follows.
flowchart TD
A[GPT-6ファミリー in GitHub Copilot] --> B[GPT-6 Astra]
A --> C[GPT-6 Sol]
A --> D[GPT-6 Luna]
C --> C1[バランス型モデル<br>インタラクティブ・エージェント型コーディング向け<br>多段階検証が必要なタスクに最適]
D --> D1[軽量・コスト効率型モデル<br>小規模・高速なタスク向け<br>GPT-6ファミリーで最安のオプション]
GPT-6 Sol
GPT-6 Sol is described as a well-rounded model designed for interactive and agentic coding. It serves as a powerful all-around choice for development tasks that benefit from careful multistep validation.
GPT-6 Luna
GPT-6 Luna is a lightweight and cost-effective model targeted at smaller, faster tasks. It is positioned as the lowest-cost option within the GPT-6 family.
Billing and Pricing Mechanism
These new models are billed under usage-based billing. Users are advised to refer to the official pricing documentation regarding GitHub Copilot models and requests for detailed pricing and unit costs per request.
Available Plans and Supported Interfaces
The primary source also clearly defines the Copilot plans that can use the models and the selectable interfaces (where the model picker is located).
Usage requirements by plan
GPT-6 Sol: Available in Copilot Pro+, Max, Business, and Enterprise plans.
GPT-6 Luna: Available in Copilot Pro, Pro+, Max, Business, and Enterprise plans (unlike Sol, the Pro plan is included).
Supported interfaces (model picker locations)
Users can select the target model from the model picker in various environments listed below.
Visual Studio Code
Visual Studio
Copilot CLI
GitHub Copilot cloud agent
GitHub Copilot app
github.com
GitHub Mobile (iOS and Android)
JetBrains
Xcode
Eclipse
Due to a phased rollout, if the model does not yet appear in the above environments, you will need to check again later.
Administrator configuration and access enablement
If organizational administrators use Copilot Business or Copilot Enterprise, model access management is handled via model policies within the Copilot settings.
Under the default model enablement settings, new models are automatically enabled unless the administrator has turned off the global default or explicitly disabled this model. Administrators can adjust these policies as needed to fit organizational requirements.
Verification Method (Prior to Device Verification)
When checking whether the models and settings covered in this article are available in your environment, verify the official information using the following steps and items.
Check whether "GPT-6 Sol" or "GPT-6 Luna" appears in the model picker within each IDE or on GitHub (may not be displayed depending on the rollout status).
Copilot Business / Enterprise administrators should check the model policy section in the settings screen.
Refer to the official documentation for detailed specifications regarding metered billing.
* Verification of GUI operations and CLI output in a Windows environment is "prior to actual device verification," and this article does not assert execution results.
Limitations
The contents of this article are based on the GitHub Changelog as of September 22, 2026, and the behavior or supported plans may change due to future specification changes or rollout progress.
Operational verification in an actual device environment has not been performed by the author, and actual screen displays and API response results depend on the reader's usage environment.
Conclusion
This article has organized official information regarding "GPT-6 Sol" and "GPT-6 Luna" added to GitHub Copilot.
Key points to check and constraints before execution:
Because the rollout is gradual, models may not appear immediately depending on the environment.
Available models vary by plan (such as the handling of Sol in the Pro plan).
Because it is based on a metered billing system, pricing details must be confirmed before use.
Administrators should pay attention to the enablement state via policy settings.
