Giving a model the ability to call external software rather than answering from memory alone. It might run a calculator for arithmetic, query a database, check today’s date, or send a request to another system. This fixes several of the things models are inherently bad at, and it is the foundation on which agents are built.
Technical view
The developer supplies structured descriptions of available functions and their parameters. When the model determines a tool is needed, it emits a structured call rather than prose; the application executes it and returns the result into the conversation for the model to use. Failure modes include calling the wrong tool, malforming arguments, and degraded selection accuracy as the number of available tools grows.
Related
Part of AI Terminology Index, section: Agents and Tools.