How to beginner · 3 min read

How to migrate from OpenAI to Claude API

Quick answer

To migrate from the OpenAI API to the Claude API, replace the OpenAI client with the Anthropic client and adjust your code to use anthropic.Anthropic with the system parameter instead of messages roles. Use client.messages.create with the Claude model like claude-3-5-sonnet-20241022 and adapt message formatting accordingly.

PREREQUISITES

Python 3.8+
OpenAI API key (for existing code)
Anthropic API key
pip install openai>=1.0
pip install anthropic>=0.20

Setup

Install both the OpenAI and Anthropic Python SDKs and set your environment variables for API keys.

Install OpenAI SDK: pip install openai>=1.0
Install Anthropic SDK: pip install anthropic>=0.20
Set environment variables:
export OPENAI_API_KEY='your_openai_key'
export ANTHROPIC_API_KEY='your_claude_key'

bash

pip install openai>=1.0 anthropic>=0.20

Step by step

Here is a complete runnable example showing how to migrate a simple chat completion from OpenAI to Claude using Python.

python

import os
from openai import OpenAI
import anthropic

# OpenAI example
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello, how are you?"}]
)
print("OpenAI response:", response.choices[0].message.content)

# Anthropic Claude example
client = anthropic.Anthropic()
message = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=300,
    system="You are a helpful assistant.",
    messages=[{"role": "user", "content": "Hello, how are you?"}]
)
print("Claude response:", message.content[0].text)

output

OpenAI response: I'm doing well, thank you! How can I assist you today?
Claude response: I'm doing great, thanks for asking! How can I help you today?

Common variations

You can customize the migration by using different Claude models, enabling streaming, or using async calls.

Use claude-3-5-haiku-20241022 for a lighter model.
Use async with asyncio and await client.messages.acreate(...).
Streaming is supported via stream=True in messages.create.

python

import asyncio
import anthropic

async def async_claude_call():
    client = anthropic.Anthropic()
    response = await client.messages.acreate(
        model="claude-3-5-sonnet-20241022",
        max_tokens=200,
        system="You are a helpful assistant.",
        messages=[{"role": "user", "content": "Tell me a joke."}]
    )
    print("Async Claude response:", response.content[0].text)

asyncio.run(async_claude_call())

output

Async Claude response: Why did the scarecrow win an award? Because he was outstanding in his field!

Troubleshooting

If you see authentication errors, verify your API keys are set correctly in environment variables. If the response is empty or incomplete, check max_tokens and model compatibility. For rate limits, implement retries with exponential backoff.

✅

Key Takeaways

Replace OpenAI client with Anthropic client and adjust message format using the system parameter.
Use Claude models like claude-3-5-sonnet-20241022 for best coding and chat performance.
Set API keys securely via environment variables for both OpenAI and Anthropic.
Async and streaming calls are supported in the Anthropic SDK for advanced use cases.
Check max_tokens and model names carefully to avoid runtime errors during migration.

Verified 2026-04 · gpt-4o, claude-3-5-sonnet-20241022, claude-3-5-haiku-20241022

Verify ↗