Tool & function calling

Models can ask to call functions you define. You pass a tools schema, the model returns tool_calls, you execute them and send the results back.

1 · Define tools and send

from openai import OpenAI
client = OpenAI(base_url="https://api.aix.theaimart.co/v1", api_key="$AIX_KEY")

tools = [{
    "type": "function",
    "function": {
        "name": "get_weather",
        "description": "Get the current weather for a city",
        "parameters": {
            "type": "object",
            "properties": {"city": {"type": "string"}},
            "required": ["city"],
        },
    },
}]

messages = [{"role": "user", "content": "What's the weather in Mumbai?"}]
resp = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V3.2",
    messages=messages,
    tools=tools,
    tool_choice="auto",
)

2 · Execute the tool calls

msg = resp.choices[0].message
if msg.tool_calls:
    messages.append(msg)                      # the assistant's tool-call turn
    for call in msg.tool_calls:
        args = json.loads(call.function.arguments)
        result = get_weather(args["city"])    # your real function
        messages.append({
            "role": "tool",
            "tool_call_id": call.id,
            "content": json.dumps(result),
        })

3 · Send results back for the final answer

final = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V3.2",
    messages=messages,
    tools=tools,
)
print(final.choices[0].message.content)

Notes

• tool_choice accepts "auto", "none", "required", or a specific function.
• The model may request several tool calls in one turn — execute all, append one tool message per tool_call_id, then call again.
• Loop steps 1–3 until the model returns content with no further tool_calls.