Implementation:Mistralai Client python Tool And Function

Knowledge Sources	Mistral Client Python Mistral AI Function Calling
Domains	Function_Calling, API_Design
Last Updated	2026-02-15 14:00 GMT

Overview

Concrete tool for defining function schemas that language models can invoke provided by the mistralai SDK's Tool and Function models.

Description

The Tool and Function Pydantic models define callable tools for the Mistral chat API. Function specifies the function's name, description, parameters (as a JSON Schema dict), and optionally strict mode for exact schema matching. Tool wraps a Function with a type field (always "function"). A list of Tool objects is passed to chat.complete(tools=...) to enable function calling.

Usage

Import Tool and Function when setting up function calling. Create one Tool per callable function with a clear description and typed parameter schema. Pass the list to chat.complete(tools=[...]) or chat.stream(tools=[...]).

Code Reference

Source Location

Repository: client-python
File: src/mistralai/client/models/tool.py (L1-20), src/mistralai/client/models/function.py (L1-24)

Signature

class Function(BaseModel):
    name: str
    description: Optional[str] = None
    parameters: Dict[str, Any]
    strict: Optional[bool] = None

class Tool(BaseModel):
    function: Function
    type: Optional[ToolTypes] = None  # Defaults to "function"

Import

from mistralai.models import Tool, Function

I/O Contract

Inputs

Name	Type	Required	Description
function.name	str	Yes	Function name for model to call
function.description	Optional[str]	No	Natural language description of function purpose
function.parameters	Dict[str, Any]	Yes	JSON Schema describing function parameters
function.strict	Optional[bool]	No	Enforce strict schema matching

Outputs

Name	Type	Description
Tool object	Tool	Validated tool definition ready for chat.complete(tools=[...])

Usage Examples

Define a Weather Lookup Tool

from mistralai.models import Tool, Function

weather_tool = Tool(
    function=Function(
        name="get_weather",
        description="Get the current weather for a given city.",
        parameters={
            "type": "object",
            "properties": {
                "city": {
                    "type": "string",
                    "description": "City name, e.g., 'Paris'"
                },
                "unit": {
                    "type": "string",
                    "enum": ["celsius", "fahrenheit"],
                    "description": "Temperature unit"
                }
            },
            "required": ["city"]
        },
    )
)

# Pass to chat.complete
response = client.chat.complete(
    model="mistral-large-latest",
    messages=[UserMessage(content="What's the weather in Paris?")],
    tools=[weather_tool],
)

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment