# Tool Calling

> Let models call your functions with OpenAI-format tool definitions.

Models marked **Tool calling** in [Models](/docs/getting-started/models) accept OpenAI-format `tools`.

```python
tools = [{
    "type": "function",
    "function": {
        "name": "get_weather",
        "description": "Current weather for a city",
        "parameters": {
            "type": "object",
            "properties": {"city": {"type": "string"}},
            "required": ["city"],
        },
    },
}]

messages = [{"role": "user", "content": "Is it raining in Singapore?"}]
first = client.chat.completions.create(model="zai-org/GLM-5.3", messages=messages, tools=tools)
call = first.choices[0].message.tool_calls[0]

messages.append(first.choices[0].message)
messages.append({"role": "tool", "tool_call_id": call.id, "content": '{"raining": true, "temp_c": 27}'})

final = client.chat.completions.create(model="zai-org/GLM-5.3", messages=messages, tools=tools)
print(final.choices[0].message.content)
```

## `tool_choice`

| Value | Behavior |
|---|---|
| `"auto"` (default) | The model decides whether to call a tool |
| `"none"` | Never call a tool |
| `"required"` | Must call at least one tool, where the model supports it |
| `{"type": "function", "function": {"name": "…"}}` | Force a specific tool, where the model supports it |

## Tips

- Keep tool descriptions short and specific; the model reads them as part of the prompt (and you pay for those tokens).
- Validate tool arguments before executing them — they're generated text.
- For many parallel calls, check `parallel_tool_calls` support per model in [models.json](/models.json).