Rate limits and retries

Throttling behavior, recommended backoff, and when to plan capacity with Yuno.

Yuno does not currently enforce per API key rate limits at the application layer. Edge infrastructure may apply per IP throttling at peak load. Clients that observe HTTP 429 Too Many Requests should back off and retry. Yuno does not currently emit X-RateLimit-Limit, X-RateLimit-Remaining, or X-RateLimit-Reset response headers, so do not rely on them in your client.

Recommended client behavior

Practice	Detail
Backoff	Exponential backoff with full jitter, starting at 1 s and capped at 30 s, on 429 and 5xx.
Retries	GET requests are safe to retry. For non idempotent writes, set a unique business key (merchant_customer_id, merchant_order_id) so retries cannot create duplicates. See Avoiding duplicates.
Retry-After	When present on a 429, honor it as the minimum wait. When absent, fall back to your jittered backoff schedule.
Self imposed ceiling	For batch jobs, target 100 requests per minute per source IP as a conservative ceiling.

If you expect traffic above that ceiling for sustained periods, contact support so the team can plan capacity with you.