What is Function Calling?

Function calling (also called 'tool use') is a feature of modern LLMs where the model can be given a list of available functions — each with a name, description, and parameter schema. When the model determines that calling a function would help complete a task, it outputs a structured call specifying which function to invoke and with what arguments. The calling application then executes the function and returns the result to the model.

This is what separates a chatbot from an agent. Without function calling, the LLM can only generate text. With it, the LLM can search the web, query a database, send an email, update a CRM record, trigger a webhook, or call any API you connect to it — and then use the results in its reasoning.

From a security standpoint, function calling requires careful design. Functions should have least-privilege access, destructive actions (like deleting records or sending emails) should have confirmation steps, and inputs should be validated before execution. Poorly designed tool sets can be manipulated via prompt injection attacks.