← Volver a las guías

Call Claude and GPT from Python with 59API

Why use a relay for Claude and GPT?

If you are building with Python, the fastest way to work with multiple AI models is to use one API pattern and switch providers behind it. 59API is an AI API relay that gives you pay-as-you-go access to Claude models like Opus, Sonnet, Haiku, and Fable, plus GPT models, through a single OpenAI-compatible endpoint: https://api.59api.com.

That matters for two reasons. First, you avoid maintaining separate integrations for different vendors. Second, you keep costs down because 59API is positioned as one of the cheapest relays while still using native official-quality models, without forcing a lower-quality downgrade. If you also want a referral rebate, that can further reduce your effective spend.

Step 1: Prepare your Python project

Start with a clean virtual environment and install the standard OpenAI SDK, because 59API is compatible with it. This is the simplest way to call both Claude and GPT from the same Python codebase.

• Python 3.9+ is a sensible baseline.
• openai is the only library you need for the API calls.
• an API key from 59API is required.

Install the SDK:

pip install openai

Step 2: Set the base URL and key

In 59API, the important difference is the base URL. Instead of pointing your client at the default OpenAI endpoint, configure it to use https://api.59api.com. Keep your key in an environment variable so it does not end up in source control.

Example environment setup:

export OPENAI_API_KEY="your_59api_key"

Step 3: Make your first GPT request

Use the OpenAI-compatible client exactly as you would for a regular OpenAI integration, but pass the 59API base URL. This example uses a GPT model name supported by your account.

from openai import OpenAI

client = OpenAI(
api_key="your_59api_key",
base_url="https://api.59api.com"
)

response = client.chat.completions.create(
model="gpt-4o-mini",
messages=[
{"role": "system", "content": "You are a concise Python tutor."},
{"role": "user", "content": "Explain list comprehensions with one example."}
]
)

print(response.choices[0].message.content)

This pattern is useful because you can keep your code stable while changing models as your needs change, whether you are optimizing for speed, quality, or price.

Step 4: Call Claude from the same Python code

The same client can call Claude models through 59API. That is the real advantage of using a relay: you do not need a separate Claude-specific SDK path just to switch providers.

response = client.chat.completions.create(
model="claude-sonnet",
messages=[
{"role": "system", "content": "You are a careful coding assistant."},
{"role": "user", "content": "Review this function for edge cases: def divide(a, b): return a / b"}
]
)

print(response.choices[0].message.content)

In practice, you can route tasks to the model that fits best. Use a cheaper model for drafting, a stronger model for difficult reasoning, and keep the same Python shape throughout your app.

Step 5: Build a small helper for model switching

Once you have both Claude and GPT working, add a helper function so your application can choose the model by task. This keeps your code readable and makes experimentation easier.

def ask_ai(model, prompt):
result = client.chat.completions.create(
model=model,
messages=[{"role": "user", "content": prompt}]
)
return result.choices[0].message.content

print(ask_ai("claude-haiku", "Summarize this paragraph in two bullets."))
print(ask_ai("gpt-4o-mini", "Rewrite this paragraph for a product spec."))

That tiny abstraction is enough for most apps. You can later add retries, logging, token tracking, or a task-based router without changing how the rest of the app calls the helper.

Step 6: Add basic production safeguards

For real applications, make the wrapper a little more robust. At minimum, handle timeouts, validate user input, and log the model name and request size. If you are sending requests from a web app or worker, keep the API key server-side only.

• Use environment variables for secrets.
• Track model usage so you know what is costing money.
• Choose smaller models first when the task does not need a premium one.
• Reuse one client instead of rebuilding it for every request.

Step 7: Sign up and start testing

If you want one Python integration for both Claude and GPT, 59API is a practical place to start. It is low-cost, pay-as-you-go, and compatible with the tools many teams already use, including Claude Code, Codex, and OpenAI SDK-based applications. Sign up, wire in the base URL, and begin by testing one simple prompt in your development environment.

Once that works, you can move on to real workloads such as chat assistants, document summarizers, coding tools, and internal automation without rewriting your model integration each time.

Precios · Docs