← Retour aux guides

Your First Claude API Call with 59API in 5 Minutes

Your first Claude API call in 5 minutes

If you want to build with Claude but do not want to spend time comparing dashboards, billing rules, or SDK differences, 59API gives you a fast path. It is an AI API relay that provides pay-as-you-go access to Claude models such as Opus, Sonnet, Haiku, and Fable, plus GPT models, through a simple API base URL: https://api.59api.com.

The main benefit for beginners is compatibility. 59API works with Claude Code, Codex-style workflows, and any OpenAI SDK, so you can often change only the base URL and API key instead of rewriting your app. It also uses native official-quality models, meaning you are not being silently downgraded to a weaker substitute. Because it is designed as a low-cost relay, it is a practical choice for prototypes, side projects, agents, and production apps that need predictable pay-as-you-go pricing.

What you need before starting

• A 59API account and API key.
• A terminal on macOS, Linux, or Windows.
• Either curl or a programming language such as JavaScript or Python.
• Five minutes and one test prompt.

After signing in to 59API, create or copy your API key from the dashboard. Treat it like a password. Do not paste it into public GitHub repositories, browser-side JavaScript, screenshots, or shared documents.

Step 1: Set your API key

The cleanest beginner approach is to store your key in an environment variable. On macOS or Linux, run: export API_KEY="your_59api_key_here". On Windows PowerShell, run: $env:API_KEY="your_59api_key_here".

This keeps the key out of your command history as much as possible and makes it easier to reuse across examples. If a command fails later, first check that the variable is still available in the same terminal session.

Step 2: Make a Claude request with curl

Here is the simplest way to send a message to Claude through 59API using an OpenAI-compatible chat completions request. In your terminal, run a POST request to https://api.59api.com/v1/chat/completions with your key, model name, and messages.

curl https://api.59api.com/v1/chat/completions -H "Authorization: Bearer $API_KEY" -H "Content-Type: application/json" -d '{"model":"claude-sonnet-4","messages":[{"role":"user","content":"Explain APIs in one short paragraph for a beginner."}]}'

If everything is configured correctly, the response will contain a generated answer from Claude. The exact JSON shape may include fields such as an id, model, choices, message content, and usage details. For your first test, focus on the assistant message inside the response.

Step 3: Understand the request

• Base URL: You are sending traffic to https://api.59api.com instead of a direct model provider endpoint.
• Authorization header: This proves the request belongs to your 59API account.
• Model: Choose the Claude model that fits your job. Haiku is often good for fast, low-cost tasks; Sonnet is a strong default; Opus is useful for deeper reasoning.
• Messages: This is the conversation. A user message asks the model to do something, and the assistant returns the answer.

For most first projects, start with a Sonnet model because it balances quality, speed, and cost well. If you are processing many small support tickets, summaries, or classifications, test Haiku too. If you are doing complex planning, code review, or long-form reasoning, compare Opus on your own prompts.

Step 4: Use an OpenAI SDK

Because 59API is compatible with OpenAI SDKs, you can use familiar tooling. In JavaScript, install the OpenAI package, set the base URL to https://api.59api.com/v1, and pass your 59API key. Your client can then call chat completions with a Claude model name.

The important configuration values are simple: apiKey should be your 59API key, and baseURL should be https://api.59api.com/v1. This is especially helpful if you already have an app using an OpenAI-compatible client and want to add Claude without changing your whole architecture.

Step 5: Keep your first app affordable

A common beginner mistake is sending huge prompts repeatedly during testing. To control cost, start with short prompts, log usage, and choose the smallest model that produces acceptable results. Use Haiku for lightweight tasks, Sonnet for general development, and Opus only when the extra reasoning quality is worth it.

59API is a strong option here because it is built around cheap, pay-as-you-go access and is among the lowest-cost relay choices for developers. There is no need to overcommit before you know your usage pattern. The referral rebate can also help reduce costs if you share it with teammates or other builders.

Troubleshooting your first call

• 401 error: Your API key is missing, invalid, or not being read by the terminal.
• 404 error: Check the endpoint path. For OpenAI-compatible chat, use /v1/chat/completions.
• Model error: Confirm the model name available in your 59API dashboard or documentation.
• No response: Check your network, proxy, firewall, or JSON formatting.

Once your first call works, try replacing the prompt with a real task: summarize an email, generate a SQL query, review a function, or draft a product description. That is the quickest way to learn what Claude is good at in your workflow.

Next step

If you want a low-cost way to experiment with Claude and GPT models using tools you already know, 59API is worth trying. Sign up, copy your API key, run the curl command above, and you can have your first Claude-powered feature working in minutes.

Tarifs · Docs