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.
You can export an API definition from your Azure API center as a connector in Microsoft Copilot Studio (preview). Copilot Studio is a graphical, low-code tool for building agents and agent flows and is part of the Microsoft Power Platform family.
In Microsoft Copilot Studio, citizen developers can use the exported connector as a custom connector to extend the capabilities of AI agents.
In this article, you create a custom connector from an OpenAPI definition associated with an API in an API center.
Prerequisites
- An API center instance. You can create one in the Azure portal. 
- An API registered in the API center inventory. The API must have a version associated with a deployment, and it must have an API definition in OpenAPI 2.0 format. 
- A Power Platform environment. 
- A Microsoft Copilot Studio license. 
Export API definition from API Center
Follow these steps to export an API definition from your Azure API center as a custom connector in Microsoft Copilot Studio.
- Navigate to your API center in the Azure portal. 
- In the left menu, select Copilot Studio (preview). 
- Select Create a connector. The Create a connector window opens. 
- Select a Power Platform environment to export the API to. 
- Enter a Connector display name. 
- Select the API, API version, and Deployment for the connector. The Host URL is populated with the deployment's runtime URL. 
- Select the API definition to use for the connector. 
- Optionally, if you configured API access settings for the API version in your API center, select an available Security requirement type to use for the connector. 
- Select Create.   - The connector is now available for use in Microsoft Copilot Studio. - After creating the connector: - You can test the connector in Power Automate or Power Apps. 
- After testing, use the connector in Microsoft Copilot Studio to create applications and workflows.