Run in production
Rate limits
Limits apply per partner, per minute, in three buckets. Every response tells you where you stand.
Buckets
Limits apply per partner (or per client IP for anonymous requests) per minute:
| Bucket | Default | Endpoints |
|---|---|---|
read | 300/min | Content, reservations and feeds: every GET except the ones below |
search | 120/min | /availability, /properties/{code}/availability, /properties/{code}/calendar |
write | 60/min | Every POST, PATCH and DELETE |
Contracted quotas enforced at Kong Gateway are separate from these limits.
Headers
Every response carries the limit and what's left of it for that bucket:
X-RateLimit-Limit: 300
X-RateLimit-Remaining: 298
When you hit a limit
You get 429 with problem type rate-limited and a Retry-After header in seconds. Wait that long before retrying.
Search less, sync more