Tutorials · Chapter D (4/4) · ~11 min

Function calling in code

Try it → see it → read → next

Parse tool calls from the model, run your functions safely, return results to the chat.

On paths: Builder · Student

Try yourself

Code Lab

Tools in code

Wire function calling: tool name → function → result → reply.

Python

# Tools in code — model chooses a tool name + args
def weather(city):
    return f"Weather in {city}: 28C sunny"

def calculator(expr):
    # tiny safe demo calculator for a few expressions
    allowed = {
        "17*24": "408",
        "17×24": "408",
        "10+5": "15",
        "100/4": "25",
    }
    return allowed.get(expr.replace(" ", ""), "0")

TOOLS = {
  "weather": weather,
  "calculator": calculator,
}

def run_tool(name, arg):
    return TOOLS[name](arg)

# Pretend the model requested a tool:
call = {"name": "calculator", "arg": "17*24"}
result = run_tool(call["name"], call["arg"])
print("tool result:", result)
print("final reply: The answer is", result)

# Challenge: change the call to weather for Dubai

Recap

What you just did

You traced model → tool name + args → your function → final natural-language reply.

Teach

How it works

Register tools with JSON schemas
Model returns {name, arguments}
Validate args → call allowlisted function
Send {role: "tool", content: result} back
Model writes user-facing answer

Same pattern across OpenAI, Anthropic, and others (names vary).

Use it

When you'd use this

Calculators, CRM lookups, booking, internal APIs
Anything live data or actions

Watch out

Never expose destructive tools without confirmation. Validate every argument.

Try next

Try this next

How-to: add function calling · Concept: structured outputs