How to beginner · 3 min read

How to define a tool for Claude API

Q: How to define a tool for Claude API

To define a tool for the Claude API, use the anthropic Python SDK by creating a client with your API key and calling client.messages.create with the system parameter to set tool behavior and messages for user input. This lets you customize Claude's responses as a tool within your application.

Quick answer

To define a tool for the Claude API, use the anthropic Python SDK by creating a client with your API key and calling client.messages.create with the system parameter to set tool behavior and messages for user input. This lets you customize Claude's responses as a tool within your application.

PREREQUISITES

Python 3.8+
Anthropic API key
pip install anthropic>=0.20

Setup

Install the Anthropic Python SDK and set your API key as an environment variable to authenticate requests.

bash

pip install anthropic>=0.20

Step by step

Define a tool by specifying the system prompt to instruct Claude on its role and use client.messages.create to send user messages. The example below shows a simple tool that answers questions about weather.

python

import os
import anthropic

client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])

system_prompt = "You are a helpful weather information tool. Answer user queries about weather concisely."
user_message = "What's the weather like in New York today?"

response = client.messages.create(
    model="claude-3-5-haiku-20241022",
    max_tokens=200,
    system=system_prompt,
    messages=[{"role": "user", "content": user_message}]
)

print(response.content[0].text)

output

The weather in New York today is mostly sunny with a high of 68°F (20°C) and a low of 54°F (12°C).

Common variations

Use different Claude models like claude-3-5-haiku-20241022 for shorter responses.
Adjust max_tokens to control response length.
Implement async calls with asyncio and client.messages.acreate for concurrency.

python

import asyncio
import os
import anthropic

async def async_tool_call():
    client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])
    system_prompt = "You are a concise assistant tool."
    user_message = "Summarize the benefits of AI tools."

    response = await client.messages.acreate(
        model="claude-3-5-haiku-20241022",
        max_tokens=100,
        system=system_prompt,
        messages=[{"role": "user", "content": user_message}]
    )
    print(response.content[0].text)

asyncio.run(async_tool_call())

output

AI tools improve efficiency, automate tasks, and enhance decision-making by providing fast, accurate insights.

Troubleshooting

If you get authentication errors, verify your ANTHROPIC_API_KEY environment variable is set correctly.
For rate limit errors, reduce request frequency or check your quota on the Anthropic dashboard.
If responses are incomplete, increase max_tokens or check for network issues.

✅

Key Takeaways

Use the system parameter to define the tool's behavior in Claude API calls.
Always authenticate with your API key from environment variables using the Anthropic SDK.
Adjust max_tokens and model choice to tailor response length and style.
Async calls with client.messages.acreate enable efficient concurrent tool usage.
Check environment variables and quotas to troubleshoot common API errors.

Verified 2026-04 · claude-3-5-haiku-20241022

Verify ↗