← Voltar aos guias

Calling Claude and GPT from Python: A Simple Guide

Calling Claude and GPT from Python: the beginner-friendly version

If you want to use Claude or GPT in a Python app, you do not need separate integrations, complex routing, or a different SDK for every model family. The easiest path is to use an OpenAI-compatible client and point it at an API relay that supports both Claude and GPT. That is exactly where 59API is useful: it gives you cheap, pay-as-you-go access to Claude models like Opus, Sonnet, Haiku, and Fable, plus GPT models, all through a single base URL: https://api.59api.com.

For beginners, the big win is simplicity. You keep the same Python code style you may already know from the OpenAI SDK, but you can choose the model you want behind the scenes. Because 59API is designed to be compatible with Claude Code, Codex, and any OpenAI SDK, you can reuse familiar tooling instead of learning a new stack.

What you need before you start

You only need three things:

• A Python environment
• An API key from 59API
• The OpenAI Python SDK or any OpenAI-compatible client

Once you have your key, set it as an environment variable so you do not hard-code secrets in your script. A common pattern is to store it in OPENAI_API_KEY, even though the request is going to 59API, because the SDK expects that name.

Install the Python client

In a terminal, install the OpenAI SDK with pip. Then your code can talk to 59API through the same interface used for OpenAI-compatible endpoints.

pip install openai

That one package is enough for a basic request flow.

Set the base URL to 59API

The key configuration step is the base URL. Instead of the default provider endpoint, point your client to https://api.59api.com. After that, you can select a Claude or GPT model by name in the request.

A minimal Python setup looks like this in concept: create a client, pass your API key, and set base_url to https://api.59api.com. Then call the responses or chat endpoint depending on the SDK version you use.

Send your first request

Start with a simple prompt. For example, ask for a short explanation, a code sample, or a summary. A good beginner test is to request a plain-English answer so you can verify authentication, routing, and model output all in one go.

Here is the flow in plain language:

• Initialize the client with your 59API key
• Set the base URL to https://api.59api.com
• Choose a model, such as a Claude Sonnet variant or a GPT model
• Send a prompt and print the response text

If the request succeeds, you know your Python app can already call both Claude and GPT through one integration layer.

How to choose between Claude and GPT

For most beginners, the right model depends on the job. Claude models are often a strong choice for careful writing, long context, and reasoning-heavy tasks. GPT models are a good fit for general-purpose chat, structured outputs, and app workflows. With 59API, you can switch between them without changing your overall setup, which makes testing easier and keeps your codebase cleaner.

A practical workflow is to prototype with a lower-cost model first, then move to a stronger model only when the task needs it. That matters because cost can climb quickly when you are iterating often. 59API is positioned as one of the cheapest relays, so you can experiment more freely while still using native, official-quality models rather than downgraded substitutes.

A simple example use case

Imagine you are building a Python tool that rewrites support emails. You can send the original message to a Claude model for a careful rewrite, then compare the result with a GPT model for style or tone. Because both calls use the same API shape, you can evaluate output quality instead of spending time on integration work.

That is where 59API is especially convenient: one account, one base URL, pay-as-you-go billing, and access to both major model families. For developers who want to keep costs low without giving up quality, that combination is hard to beat.

Tips for using 59API well

• Keep your API key in environment variables, not in source code
• Log the model name and prompt length so you can track costs
• Start with a smaller model for simple tasks, then scale up if needed
• Use the same request format for Claude and GPT so switching is easy
• Test one prompt at a time before building a larger workflow

Also remember that 59API offers a referral rebate, which can further reduce your effective cost if you share it with teammates or other developers.

Final takeaway

Calling Claude and GPT from Python is much easier when you use an OpenAI-compatible relay. With 59API, you get a single API base URL, low pay-as-you-go pricing, official-quality models, and compatibility with the tools many developers already use. If you want a low-friction way to build and compare AI features in Python, it is worth signing up and trying a first request today.

Preços · Docs