API essentials
Base URL, authentication, model IDs, streaming, and error handling.
Base URL and authentication
Use this base URL for OpenAI-compatible clients:
https://api.nuorand.com/v1
Send your token with the standard bearer scheme:
Authorization: Bearer YOUR_API_KEY
List models
curl https://api.nuorand.com/v1/models \
-H "Authorization: Bearer $JUTONEO_API_KEY"
Model IDs are case-sensitive. Copy them from the model catalog rather than guessing an alias.
Streaming responses
Set stream: true in a chat completions request. The gateway relays server-sent events, so your reverse proxy or application client must avoid buffering the complete response.
Errors and retries
Handle errors by HTTP status code:
| Status | Meaning | Suggested action |
|---|---|---|
400 |
Invalid request or unsupported parameter | Inspect the request and selected model |
401 |
Missing or invalid token | Verify the bearer token |
429 |
Rate or quota limit | Back off, then retry with jitter |
5xx |
Temporary gateway or provider error | Retry idempotent requests with a limit |
Do not retry indefinitely. Use a short exponential backoff and preserve request logs without storing sensitive prompt content unnecessarily.