1. Create your key
Save the full key when you create it. Keep it in a server-side environment variable.
Manage API keys →Managed AI Gateway
One account for credits, API keys and usage. You do not need an upstream provider key to use managed access.
Save the full key when you create it. Keep it in a server-side environment variable.
Manage API keys →Use an exact model ID from the catalog. Availability and billing depend on your account and selected model.
Browse models →Start with a short request, then check the model, usage and balance in your account.
View usage →Base URL: https://api.ai24x.com/v1
Set AI24X_API_KEY on your server. This example uses a POSIX shell.
curl https://api.ai24x.com/v1/chat/completions \
-H "Authorization: Bearer $AI24X_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"auto","messages":[{"role":"user","content":"Hello"}],"max_tokens":128}'On Windows, use the PowerShell guide. For applications, see Node.js or Python / HTTP.
Send Authorization: Bearer followed by your complete AI24X platform key. A login session token is not the key to paste into an external tool.
Set stream to true for incremental responses. Test the non-streaming request first when diagnosing a connection problem.
Use the model field in each request. A model listed in the catalog may have account, balance or provider-specific restrictions.
Check that the key is complete and active. Replace a revoked or exposed key in API keys.
Check your balance and the selected model. A BYOK service subscription and managed credits are different products.
Reduce concurrency and wait before retrying. Avoid immediate retry loops.
Check service status. Retry with a short request and record the time, model and request ID, but never send your key to support.