Step 11 of 15
TL;DR: Tool use (also called function calling) lets an AI use real tools — search the web, send an email, run code, check a database. It's how AI stops just talking and starts doing.
Imagine a super-smart brain floating in a jar. It knows a lot, but it has no hands, so it can't do anything. Tool use gives that brain a set of hands and buttons it can press. Now it can actually get stuff done.
Tool use is when you give the AI a list of actions it's allowed to take — like "search," "send email," or "look up an order" — and let it choose which to use. The model decides when a tool is needed and asks to run it.
On its own, an LLM only knows old, general info and can only write text. With tools it can get fresh data, do exact math, and change things in the real world. That's the leap from a talker to a worker.
Remember agents from the last step? Tool use is their hands. An agent's "act" step is almost always a tool call. Master tool use and you can build AI that truly does things.
Yes, they mean the same thing. "Function calling" is the technical name; "tool use" is the friendly one. Both mean letting the AI trigger real actions.
No. The AI only asks to run a tool. Your code actually runs it and returns the result. This keeps you in control of what's allowed.
Almost anything you can write code for: web search, databases, email, calculators, calendars, internal company systems, and more.
Want to know exactly where you stand on the road from using AI to building it? Take the free 3-minute AI Builder Scorecard and get your single biggest gap.