In ParadeDB, a “term” is equivalent to a “token.” A token is a chunk of text that can be indexed and searched. Typically, tokens are individual words, but they can also be smaller fragments, depending on the tokenizer being used.