Online testing

Before production integration, use the built-in tester on this page to run a small request and verify the key, model name, Base URL, and balance.

• Create an API key and copy an available model name from the console.
• Enter the API key in the tester below, and change the default model name if your console shows a different model.
• Send the request and confirm the response panel returns a status code and model output.
• If you later use ChatBox, Cherry Studio, Postman, or curl, reuse the Base URL and request body shown in the documentation.
• Prefer OpenAI Compatible, Custom OpenAI, or OpenAI API Compatible Provider mode when available.
• Do not run high-frequency health checks against large models; they consume credits and may trigger rate limits.