Model availability and controls (GitHub Copilot)

Microsoft Copilot Studio supports multiple generative AI models with different capabilities, performance characteristics, and availability requirements. You can select a model for an agent that uses the GitHub Copilot harness. You can also explore experimental and preview models before they become generally available.

This article explains model release types, currently listed models, retirement dates, regional availability, and administrator controls. Model availability can vary by region, environment settings, and release stage.

Model release types and availability labels

Release types describe production readiness. Lifecycle and availability labels describe a model's status or conditions for use.

Release type

Other label

Model lifecycle summary

The Available models table shows the most recent model changes. The Retirement table identifies models with announced lifecycle changes. For geographic restrictions, see Model availability by region.

Available models

Model retirement lifecycle

The following models are retiring soon or have retired. Review the listed replacement and any applicable exceptions. A replacement designation doesn't by itself establish whether migration is automatic.

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.

Model use categories

The Auto category supports mixed workloads by routing requests dynamically, including knowledge and action scenarios. Latency, cost, and reasoning effort might vary by request.

External models

Copilot Studio also supports external model offerings, including Anthropic and Mistral models. Availability and requirements depend on the model and applicable terms.

• 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.

Limitations of experimental and preview models

Experimental and preview models are intended for evaluation and testing, not production use. They might vary in response quality, latency, performance, and message consumption. They might time out or become unavailable. These models are subject to preview terms.

If you publish an agent that uses an experimental or preview model, usage is billed at the applicable established rates. Review current billing guidance before publication.

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.

For models marked cross-geo, administrators should review data movement requirements and the organization's geographic restrictions before enabling access.

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:

  • Connect to Anthropic LLM

  • Connect to Mistral

  • Connect to xAI

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.

External models might require extra administrator approval and acceptance of provider terms. For guidance on the specific provider and model, review the current external-model guidance.

Related information

• Select an agent model
• Preview terms
• Early release cycle environments
• Preview and experimental AI models
• Geographical availability and data movement
• Billing and credits