Function Calling: Let an LLM Use Your Tools

The idea

Function calling lets you describe tools (as JSON schemas) to the model. Instead of answering directly, the model can return a structured request to call one of your functions — you run it and feed the result back.

tools = [{
  "type": "function",
  "function": {
    "name": "get_weather",
    "description": "Get current weather for a city",
    "parameters": {"type": "object",
      "properties": {"city": {"type": "string"}}, "required": ["city"]},
  },
}]
resp = client.chat.completions.create(model="gpt-4o-mini",
  messages=[{"role":"user","content":"Weather in Paris?"}], tools=tools)

The loop

Call the model → if it requests a tool, run it → append the result as a tool message → call again. This loop is the foundation of every AI agent.

Function Calling: Let an LLM Use Your Tools

What You'll Learn

Tutorial Content

The idea

The loop

Your Progress

Tags