Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Note
The harness used to build this agent uses usage-based billing. Using, building, testing, and evaluating agents might consume Copilot Credits. Learn more in Overview of usage-based billing.
AI capabilities evolve rapidly, and each generative model brings distinct strengths, whether it's faster responses, higher-quality outputs, or improved cost efficiency. By using Copilot Studio, you can choose the best model for your agent's orchestration by using a simple dropdown menu.
Want to try out cutting-edge models before they're production-ready? Access the latest experimental models to evaluate them early. However, they might have limited testing, availability, and functionality.
This article describes the agents available for agents and workflows powered by the GitHub Copilot harness.
Model lifecycle summary
Use this summary as the primary lifecycle view. For regional details, see Model availability by region.
Available models
| Model | Release type |
|---|---|
| GPT-5 Chat | Generally available (GA, retiring soon) |
| GPT-5.6 Sol | Generally available (GA, retiring soon) |
| GPT-5.6 Luna | Generally available (GA, retiring soon) |
| GPT-6 Astra | Generally available (GA) |
| GPT-6.1 Sol | Generally available (GA) |
| GPT-6 Luna | Generally available (GA) |
| Claude Sonnet 4.6 | Generally available (GA) |
| Claude Sonnet 5 | Generally available (GA) |
| Claude Fable 5 | Generally available (GA) |
| Claude Fable 5.1 | Generally available (GA) |
| Claude Opus 4.8 | Generally available (GA) |
| Claude Opus 5 | Generally available (GA) |
| Mistral Medium 3.5 | Experimental |
Model retirement dates and replacements
| Retiring model | Retirement date | Replacement model |
|---|---|---|
| GPT-5.5 Chat | October 31, 2026 | GPT-6 Luna |
| GPT-5.6 Sol | November 12, 2026 | GPT-6.1 Sol |
| GPT-5.6 Luna | November 12, 2026 | GPT-6 Luna |
| GPT-5 Chat | November 12, 2026 | GPT-6 Luna |
Note
GPT-5 Chat retires on November 12, 2026, except for FedRAMP customers and customers outside the United States whose environments restrict cross-geo data movement. GPT-5 Chat remains available to these customers until further notice.
Model availability by region
The following table shows availability status across regions and special scopes. Availability is listed as available by default. A - means the model is explicitly unavailable in that region based on model-definition rules.
GitHub Copilot harness availability
| Model | Asia | Australia | Brazil | Canada | Europe (Except UK) | India | Japan | Korea | Saudi Arabia | Singapore | South Africa | United Kingdom | United States |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GPT-5 Chat | GA (cross-geo) | GA | GA (cross-geo) | GA (cross-geo) | GA | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA | GA |
| GPT-5.5 Chat | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA |
| GPT-5.6 Sol | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) |
| GPT-5.6 Luna | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) |
| GPT-6 Astra | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA |
| GPT-6.1 Sol | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) |
| GPT-6 Luna | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) |
| Claude Sonnet 4.6 | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA |
| Claude Sonnet 5 | GA (cross-geo) | - | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | - | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (early access environment) |
| Claude Fable 5 | GA (cross-geo) | - | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | - | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (early access environment) |
| Claude Fable 5.1 | GA (cross-geo) | - | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | - | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (early access environment) |
| Claude Opus 4.8 | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA |
| Claude Opus 5 | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA (cross-geo) | GA |
| Mistral Medium 3.5 | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) | Experimental | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) | Experimental (cross-geo) |
Note
Cross-geo behavior applies to this availability table. A model can be listed as available in a region even when processing is routed outside that region. If your environment doesn't permit cross-geo data movement, that model might be unavailable at runtime. Learn more in Admin controls for AI model selection.
Model use categories
Models are optimized for different purposes. The following table describes the model use category available for this GitHub Copilot harness view.
| Tag | Description | Strengths | Latency | Cost | Reasoning depth |
|---|---|---|---|---|---|
| Auto | Optimized for coverage across mixed workloads; routes queries dynamically. | Helpdesk and employee agents with mixed intents, blending knowledge and actions, and tier‑0 customer support. | Variable | Variable | Adaptive per turn |
Model release types
Each model listed in Copilot Studio has a tag that identifies its release type. You can try new, cutting-edge experimental and preview models, or choose a reliable, thoroughly tested generally available model.
- Experimental: Used for experimentation, and not intended for production use. Subject to preview terms, and can have limitations on availability and quality. See Limitations of experimental and preview models.
- Preview: Might eventually become a generally available model, but currently not intended for production use. Subject to preview terms, and can have limitations on availability and quality. See Limitations of experimental and preview models.
- Generally available (GA): Models without a release tag are generally available. You can use this model for scaled and production use. In most cases, generally available models have no limitations on availability and quality, but some might still have some limitations, like regional availability.
- Default: The default model for all agents, and usually the best-performing generally available model. The default model is periodically upgraded as new, more capable models become generally available. Agents also use the default model as a fallback if a selected model is turned off or unavailable.
- Retired: When a new model becomes the default model, the old default model is retired.
- Cross-geo: Might require data processing and storage outside of your organization's geographical boundaries. Your admin can turn data movement across regions on or off.
- Early access environment: An early release cycle environment. These environments receive new updates and features first. Learn more in Early release cycle environments.
External models
You can also add external AI models from Anthropic or Mistral to your agent.
Copilot Studio offers models from multiple providers for your agents. Read the list of models available in Copilot Studio.
Anthropic: Anthropic is a Microsoft subprocessor that operates with Microsoft oversight through contractual safeguards and appropriate technical and organizational measures. Learn more about subprocessors.
Mistral: Mistral models are hosted outside Microsoft and are subject to Mistral terms and data handling. Admins need to review and accept these terms before makers can use the models. Learn more about using Mistral models. These models are available before an official release so that you can get early access and provide feedback.
Important
Don't use external models tagged preview or experimental for production purposes. You might experience slowdowns or timeouts due to limited capacity and availability, and these models might not be supported in the future. Admins can control access to this feature. Experimental and preview models are subject to preview terms. These models are available before an official release so that you can get early access and provide feedback.
This article contains Microsoft Copilot Studio external model documentation and is subject to change.
Limitations of experimental and preview models
Important
- Experimental models are available for exploration and testing but aren't recommended for production use. Review Limitations of experimental and preview models before choosing an experimental or preview model for your agent.
- Data processed within an experimental model might be processed and stored outside of your organization's geographical boundaries.
- This article contains Copilot Studio documentation on model selection, which includes experimental model previews, and is subject to change.
You can explore and test experimental and preview models, but don't use them for production:
They might show variability in performance, response quality, latency, or message consumption, and might time out or be unavailable.
If you publish an agent with an experimental or preview model, and users use the agent, that use is billed at the established rates.
Feel free to experiment with these models. However, be cautious about deploying them in production environments.
Experimental and preview models are subject to preview terms. Microsoft makes these models available before an official release so that you can get early access and provide feedback.
Admin controls for AI model selection
Administrators can use the following settings to allow or block makers from adding preview and experimental AI models to agents:
Administrators can choose to allow or block preview and experimental models in an environment. To use these models, turn on the Preview and experimental AI models setting for your environment.
Data processed within a preview or experimental model might be processed and stored outside of your organization's geographical boundaries. To make experimental models available, your environment must have the Move data across regions setting turned on. The tenant administrator manages this environment-level setting in the Power Platform admin center.
Admin controls and requirements for external models
Admins control whether makers can add external models to agents. To grant access to external models, admins must complete the following actions:
Turn on external models in Power Platform admin center for the environment or the environment group.
Allow access to each external model provider in the Microsoft 365 admin center. Learn more in the Microsoft 365 admin center documentation:
Preview models and external models are two different sets that can overlap but aren't the same, and their settings are separate. For example:
Admins can block external models but allow preview or experimental models. In this case, makers can't use external models but can use preview, experimental, and generally available internal models.
Admins can block preview or experimental models but allow external models. In this case, makers can't use any preview or experimental models, but can use any generally available external and internal models.