A model that accepts text as input and produces text as output, without support for images, audio, or other data types.