Tool Use / Function Calling
Tool use, also called function calling, is how an AI model goes beyond generating text to take action, by calling external functions, APIs, or data sources to fetch information or do something.
It lets a model look up live information or take an action, such as querying a database or sending an email, instead of only answering from memory. On its own, a language model can only produce text based on its training; function calling gives it a defined set of tools it can invoke, passing structured inputs and receiving results it can act on.
This is the mechanism that turns a model into an agent that can retrieve current data and change the state of other systems. It underpins agentic workflows and is the foundation that newer connection standards between models and tools aim to standardize.
Frequently Asked Questions
It is a model's ability to invoke predefined functions or tools, passing structured inputs and using the results, so it can fetch data or take actions rather than only generating text.