GPT-5.6 Luna Quickstart — API Tutorial and First Request
This GPT-5.6 Luna quickstart shows the fastest production path: create one 97AI.PRO API key, call `https://97ai.97claude.com/api/chat`, verify the response, then move the request into your app. Pricing is $0.0735 per per million tokens, failed chat completions cost $0, and no OpenAI account is required.
1-minute setup
- Create a 97AI.PRO account and open the API keys page.
- Top up at least $5 by credit card, USDC on Arbitrum, Alipay, or WeChat Pay.
- Copy an API key and call `https://97ai.97claude.com/api/chat` with `model: "chat/gpt-5-6-luna"`.
- Log the returned credits and balance fields so your production app can show usage clearly.
Copy-ready first request
Run this from a terminal after replacing $YOUR_KEY with your live API key.
curl -X POST https://97ai.97claude.com/api/chat \
-H "Authorization: Bearer $YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"chat/gpt-5-6-luna","messages":[{"role":"user","content":"Write a short product brief"}]}'
Production checklist
- Store the API key server-side only; never expose it in browser JavaScript.
- Add a request id to your own logs before every chat completion so retries are traceable.
- Handle failed calls as a normal state. 97AI.PRO charges $0 for failures, but your UI should still show the error.
- Keep client timeouts conservative because GPT-5.6 Luna returns synchronously.
Frequently asked questions
Do I need a OpenAI account for this GPT-5.6 Luna tutorial?+
No. 97AI.PRO gives you one API key that works across GPT-5.6 Luna and the rest of the model catalog.
Which endpoint should I use for GPT-5.6 Luna?+
Use `https://97ai.97claude.com/api/chat` with `model: "chat/gpt-5-6-luna"`.
How much should I top up before testing?+
$5 is enough for a practical smoke test. The exact number of calls depends on the model and payload size.
Can I use this quickstart in China or without a US card?+
Yes. 97AI.PRO supports Alipay, WeChat Pay, and USDC on Arbitrum, and the platform is reachable from mainland China without a VPN.
What is the first production risk to test?+
Test failure handling first. Treat a failed chat completion as a normal API state, log it, show a clear message, and retry only when your app can do so safely.
Ready to try GPT-5.6 Luna?
Sign up on 97AI.PRO, top up with credit card / USDC / Alipay, and get an API key in under 60 seconds.
