We demonstrate prototypes of speech and language interfaces for intelligent agents. We deal with simulated agents operating in a virtual world as well as with physical agents in the form of remote controlled devices. The research underlying these prototype demonstrations involves various aspects of Artificial Intelligence research, in particular Natural Language and Speech Processing, and Knowledge Representation and Reasoning for Actions. The ultimate goal of this research is to develop adaptable speech based interfaces for a variety of agent-type systems, i.e. physical systems like household appliances, e.g. TVs, stereos, door openers, or alarm systems, as well as software systems with command interfaces, e.g. design systems or help systems.