← Lessons

quiz vs the machine

Gold1400

Machine Learning

Tool Use And Function Calling

Letting a model request external functions so it can act beyond text.

5 min read · core · beat Gold to climb

What it is

Tool use, often called function calling, lets a language model request that an external function run on its behalf. The model cannot search the web or query a database itself, so it emits a structured request and the surrounding system executes it.

The loop

The interaction follows a clear cycle.

  • The developer describes available tools, each with a name and an input schema
  • When the model decides a tool is needed, it outputs a structured call with arguments
  • The application runs the tool and feeds the result back into the conversation
  • The model reads the result and continues, possibly calling more tools

This separation keeps the model focused on deciding what to do, while real code does the doing.

Why it matters

Function calling turns a text predictor into an actor.

  • It grounds answers in fresh, real data instead of memorized text
  • It lets the model take actions like sending an email or booking a slot
  • The schema makes the model output machine readable, so it plugs into normal code

Key idea

Tool use lets a model emit structured function calls that the surrounding system runs, grounding answers and enabling real actions.

Check yourself

Answer to earn rating on the learn ladder.

1. What does the model actually produce when it uses a tool?

2. Why does function calling help ground answers?