Which Azure resource can host a trained Custom Vision model?

The correct answer is that a trained Custom Vision model can be hosted within Cognitive Services. This is because Azure Cognitive Services provides a set of APIs and tools that enable developers to integrate AI capabilities into their applications without needing to build machine learning models from scratch. Custom Vision is a part of Cognitive Services, specifically designed for image classification tasks.

When you train a Custom Vision model, it generates a model that can be deployed as a web service. Hosting it within Cognitive Services allows you to easily access the model via REST APIs, enabling integration into applications for real-time predictions. By leveraging Cognitive Services for deployment, users can take advantage of features such as scaling, reliability, and security, which are essential for production scenarios.

Other choices, such as Custom Vision (Training), are specific to the training aspect and may not provide the hosting capabilities needed for accessing the model once it is trained. Computer Vision focuses on broader image analysis tasks but does not specialize in Custom Vision models specifically. Machine Learning Studio is more about building and experimenting with machine learning workflows rather than the specific hosting of trained models. Thus, Cognitive Services is the ideal choice for hosting trained Custom Vision models within the Azure ecosystem.