Implementation:Cohere ai Cohere python ToolV2Function Model

Knowledge Sources	Cohere Python SDK
Domains	SDK, Tool Use, Function Calling, Pydantic Models
Last Updated	2026-02-15 14:00 GMT

Overview

Pydantic model representing the function specification within a V2 tool definition, including name, description, and JSON Schema parameters.

Description

ToolV2Function defines the function metadata for a V2 tool. It contains a required name field identifying the function, an optional description field explaining what the function does, and a required parameters field that holds a JSON Schema dictionary describing the function's input parameters. This model is referenced by ToolV2 (which wraps it under the function field with type="function") and is the primary way to declare tool capabilities when calling V2Client.chat() or V2Client.chat_stream(). The class extends UncheckedBaseModel with extra="allow" for forward compatibility.

Usage

Use ToolV2Function when building tool definitions for the Cohere V2 chat API. Each tool passed to the tools parameter of client.chat() should be a ToolV2 instance containing a ToolV2Function that describes the callable function's name, purpose, and parameter schema. The parameters field follows standard JSON Schema conventions (with "type", "properties", and "required" keys).

Code Reference

Source Location

• Repository: Cohere Python SDK
• File: src/cohere/types/tool_v2function.py Lines L1-36

Signature

class ToolV2Function(UncheckedBaseModel):
    """The function to be executed."""

    name: str = pydantic.Field()
    description: typing.Optional[str] = pydantic.Field(default=None)
    parameters: typing.Dict[str, typing.Any] = pydantic.Field()

Import

from cohere import ToolV2Function

I/O Contract

Fields

Field	Type	Required	Default	Description
name	str	Yes	(none)	The name of the function.
description	Optional[str]	No	None	The description of the function.
parameters	Dict[str, Any]	Yes	(none)	The parameters of the function as a JSON schema.

Configuration

Setting	Value	Description
extra	"allow"	Permits additional fields not defined in the schema, ensuring forward compatibility.
smart_union	True	Pydantic V1 only: enables smart union type resolution.

Usage Examples

from cohere import ToolV2, ToolV2Function

# Define a tool with a ToolV2Function specification
tools = [
    ToolV2(
        function=ToolV2Function(
            name="get_weather",
            description="Get the current weather for a given location",
            parameters={
                "type": "object",
                "properties": {
                    "location": {
                        "type": "string",
                        "description": "The city and state, e.g. San Francisco, CA",
                    },
                    "unit": {
                        "type": "string",
                        "enum": ["celsius", "fahrenheit"],
                        "description": "Temperature unit",
                    },
                },
                "required": ["location"],
            },
        )
    ),
]

# Pass tools to V2 chat
response = client.chat(
    model="command-a-03-2025",
    messages=[{"role": "user", "content": "What is the weather in Boston?"}],
    tools=tools,
)

Related Pages

• Implementation:Cohere_ai_Cohere_python_ToolV2_Definition -- Parent ToolV2 model that wraps ToolV2Function
• Implementation:Cohere_ai_Cohere_python_ToolCallV2_Model -- Tool call response model returned when the model invokes a tool
• Implementation:Cohere_ai_Cohere_python_ToolParameterDefinitionsValue_Model -- Legacy V1 parameter definition model
• Implementation:Cohere_ai_Cohere_python_Tool_Execution_Pattern -- End-to-end tool execution pattern
• Environment:Cohere_ai_Cohere_python_Python_SDK_Runtime