Code beginner · 3 min read

How to handle Claude tool call results in python

Q: How to handle Claude tool call results in python

Use the Anthropic Python SDK's client.messages.create method to call Claude models and handle results by accessing response.content[0].text for the tool call output.

Direct answer

Use the Anthropic Python SDK's client.messages.create method to call Claude models and handle results by accessing response.content[0].text for the tool call output.

Setup

Install

bash

pip install anthropic

Env vars

ANTHROPIC_API_KEY

Imports

python

import anthropic
import os

Examples

inUser asks: 'What is the capital of France?'

outResponse text: 'The capital of France is Paris.'

inUser asks: 'Summarize the latest AI trends.'

outResponse text: 'Recent AI trends include large multimodal models, improved reasoning, and efficient fine-tuning techniques.'

inUser asks: 'Translate "Hello" to Spanish.'

outResponse text: 'Hola'

Integration steps

Initialize the Anthropic client with the API key from os.environ
Prepare the messages list with the user prompt
Call client.messages.create with the Claude model and messages
Receive the response object containing the tool call results
Extract the text result from response.content[0].text
Use or display the extracted result as needed

Full code

python

import anthropic
import os

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

user_prompt = "What is the tallest mountain in the world?"

response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=500,
    system="You are a helpful assistant.",
    messages=[{"role": "user", "content": user_prompt}]
)

result_text = response.content[0].text
print("Claude response:", result_text)

output

Claude response: The tallest mountain in the world is Mount Everest, which stands at 8,848 meters (29,029 feet) above sea level.

API trace

Request

json

{"model": "claude-3-5-sonnet-20241022", "max_tokens": 500, "system": "You are a helpful assistant.", "messages": [{"role": "user", "content": "What is the tallest mountain in the world?"}]}

Response

json

{"id": "chatcmpl-xxx", "object": "chat.completion", "model": "claude-3-5-sonnet-20241022", "choices": [{"index": 0, "message": {"role": "assistant", "content": [{"type": "text", "text": "The tallest mountain in the world is Mount Everest, which stands at 8,848 meters (29,029 feet) above sea level."}]} }], "usage": {"prompt_tokens": 20, "completion_tokens": 30, "total_tokens": 50}}

Extractresponse.content[0].text

Variants

Streaming Claude Tool Call ›

Use streaming when you want to display partial results as they arrive for better user experience with long responses.

python

import anthropic
import os

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

user_prompt = "Explain quantum computing in simple terms."

# Streaming example
for chunk in client.messages.stream(
    model="claude-3-5-sonnet-20241022",
    max_tokens=500,
    system="You are a helpful assistant.",
    messages=[{"role": "user", "content": user_prompt}]
):
    print(chunk.text, end='', flush=True)
print()

Async Claude Tool Call ›

Use async calls to handle multiple concurrent Claude requests efficiently in event-driven Python applications.

python

import anthropic
import os
import asyncio

async def async_claude_call():
    client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])
    user_prompt = "List three benefits of AI in healthcare."
    response = await client.messages.acreate(
        model="claude-3-5-sonnet-20241022",
        max_tokens=200,
        system="You are a helpful assistant.",
        messages=[{"role": "user", "content": user_prompt}]
    )
    print("Async Claude response:", response.content[0].text)

asyncio.run(async_claude_call())

Use a Smaller Claude Model for Cost Efficiency ›

Use smaller Claude models like claude-3-opus-20240229 to reduce cost and latency for simpler tasks.

python

import anthropic
import os

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

user_prompt = "Give me a brief summary of climate change."

response = client.messages.create(
    model="claude-3-opus-20240229",
    max_tokens=300,
    system="You are a helpful assistant.",
    messages=[{"role": "user", "content": user_prompt}]
)

print("Claude smaller model response:", response.content[0].text)

Performance

Latency~700ms for typical Claude-3.5-sonnet calls (non-streaming)

Cost~$0.003 per 500 tokens for Claude-3.5-sonnet

Rate limitsTier 1: 300 RPM / 20K TPM

Use concise prompts to reduce token usage
Limit <code>max_tokens</code> to expected response size
Reuse context efficiently to avoid repeating information

Approach	Latency	Cost/call	Best for
Standard Claude call	~700ms	~$0.003/500 tokens	General purpose, high-quality responses
Streaming Claude call	Starts ~300ms, streams over time	~$0.003/500 tokens	Long responses with better UX
Async Claude call	~700ms per call, concurrent	~$0.003/500 tokens	High throughput, concurrent requests
Smaller Claude model	~400ms	~$0.0015/500 tokens	Cost-sensitive or simpler tasks

✓

Quick tip

Always extract the tool call result from <code>response.content[0].text</code> when using the Anthropic SDK for Claude calls.

⚠

Common mistake

Beginners often forget to use the <code>system</code> parameter or misread the nested <code>content[0].text</code> structure, causing extraction errors.

Verified 2026-04 · claude-3-5-sonnet-20241022, claude-3-opus-20240229

Verify ↗