Connect to your deployed model

Open your customer portal, confirm a running deployment and create a bounded API key. Copy the actual API base URL from Infrastructure. Keep the secret outside your source code.

# Set these locally from your actual deployment; never commit a key.
# RHEIN_API_BASE_URL includes /v1 and must use HTTPS.
curl --fail-with-body "$RHEIN_API_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $RHEIN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"customer-model","messages":[{"role":"user","content":"Hello"}],"max_tokens":128}'

401: check your key. 429: inspect its rate/token limits and budget. Unavailable service: check deployment status and contact support. A model response is never simulated by this portal.