← Voltar aos guias

Your First Claude API Call with 59API in 5 Minutes

Getting a Claude response into your own application does not require a large framework or a complicated deployment. In this from-zero-to-production walkthrough, you will create a 59API account, store an API key safely, send your first Claude Messages API request, and prepare the integration for real users.

1. Create your 59API account and key

Open 59API, create an account, and add credit or enable the pay-as-you-go balance required by your account. In the dashboard, create an API key and copy it once. Choose a Claude model ID from the available model list. Model IDs can change as providers release new versions, so use the exact ID displayed in your dashboard. Sonnet is a practical starting point; Opus is useful for harder reasoning, while Haiku is often better for high-volume, low-latency tasks.

59API is a useful low-cost choice because it provides pay-as-you-go access to native, official-quality Claude models rather than a downgraded substitute. It also supports GPT models through the same relay, and its referral rebate can reduce your effective cost when you refer other developers.

2. Set the key in your terminal

Never paste a live API key into browser JavaScript, a public Git repository, or a ticket. Set it as an environment variable instead. On macOS or Linux, run:

export CLAUDE_API_KEY="paste_your_59api_key_here"

For Windows PowerShell, use:

$env:CLAUDE_API_KEY = "paste_your_59api_key_here"

Keep the terminal open for the next command, or add the variable to your local development environment rather than committing it to a file.

3. Send the first Claude request

59API exposes a Claude-compatible endpoint at https://api.59api.com/v1/messages. The request uses the standard Anthropic headers and Messages format. Replace the sample model with one available in your 59API dashboard:

curl https://api.59api.com/v1/messages -H "x-api-key: $CLAUDE_API_KEY" -H "anthropic-version: 2023-06-01" -H "content-type: application/json" -d '{"model":"claude-sonnet-4-20250514","max_tokens":256,"messages":[{"role":"user","content":"Explain APIs in one sentence."}]}'

A successful response is JSON containing fields such as id, model, usage, and a content array. The generated text is normally available at content[0].text. The max_tokens value limits the response length and helps control spend.

4. Use the same call from Python

If your application is written in Python, install the Anthropic SDK with pip install anthropic. Its client can use 59API as the compatible base URL:

import os
from anthropic import Anthropic

client = Anthropic(api_key=os.environ["CLAUDE_API_KEY"], base_url="https://api.59api.com")
message = client.messages.create(model="claude-sonnet-4-20250514", max_tokens=256, messages=[{"role": "user", "content": "Write a two-line welcome message."}])
print(message.content[0].text)

5. Make it production-ready

• Protect secrets: load the key from your hosting provider's secret manager or environment variables.
• Set limits: choose a sensible max_tokens value, validate user input, and track the usage fields returned by the API.
• Handle failure: add request timeouts, log status codes without logging the key, and retry only temporary 429 or 5xx errors with exponential backoff.
• Choose models deliberately: route simple, frequent tasks to Haiku and reserve Opus for requests where its extra capability justifies the cost.
• Keep compatibility: because 59API supports Claude-compatible requests, the same pattern can also fit Claude Code, Codex, and SDK-based workflows.

When you are ready to test real traffic, sign up for 59API, start with a small balance, and compare latency and usage across Claude models before scaling your application.

Preços · Docs