Supports voice input/output overview

The tool or agent supports voice as both an input and output modality. Can accept spoken input (via microphone/phone) and produce voice output (via text-to-speech or direct audio). Distinct from vision-input which handles image modality.