An agent runs a model in a loop: it reasons about a goal, takes an action such as a tool call, observes the result, and repeats until the task is done. Yao et al. (2023) formalised the interleaving of reasoning and acting in the ReAct framework, and Schick et al. (2023) showed in Toolformer that models can learn to decide when and how to call external tools.
The difference from a chatbot is autonomy over several steps — which makes agents powerful, but also raises reliability and safety questions.