Differentiate Reasoning vs. non-Reasoning models in API

It would be great to have a way to determine if a model is a Reasoning model or a non-Reasoning model. This allows e.g. to add a switch to custom chatbots to limit the model selection to one type.